https://wiki.swarma.org/api.php?action=feedcontributions&feedformat=atom&user=Shenky20集智百科 - 复杂系统|人工智能|复杂科学|复杂网络|自组织 - 用户贡献 [zh-cn]2026-11-04T11:09:53Z用户贡献MediaWiki 1.35.0https://wiki.swarma.org/index.php?title=%E7%A5%9E%E7%BB%8F%E5%BD%A2%E6%80%81%E8%AE%A1%E7%AE%97&diff=33256神经形态计算2022-07-29T12:42:21Z<p>Shenky20：</p>
<hr />
<div>{{#seo:<br />
|keywords=人造神经元，神经形态计算，类脑计算<br />
|description=指使用包含电子模拟电路的超大规模集成电路系统来模拟神经系统中生理结构的研究方法<br />
}}<br />
<br />
<br />
本文主要参考并翻译的文章为"2022 roadmap on neuromorphic computing and engineering."，[48]结合维基百科词条Neuromorphic engineering[53]。<br />
<br />
==概述==<br />
<br />
神经形态工程 Neuromorphic engineering（也称为神经形态计算 Neuromorphic computing或类脑计算 Brain-inspired computing）[1][2][3]是指使用包含电子模拟电路的超大规模集成电路系统来模拟神经系统中生理结构的研究方法。神经形态计算机或神经形态芯片包括任何使用由硅制成的人造神经元进行计算的设备。[4][5]近年来，神经形态学 neuromorphic被用来描述能够实现神经系统模型功能（如感知、运动控制，多感官整合等）的模拟、数字、模拟/数字混合模式超大规模集成电路和软件系统。神经形态计算的硬件实现可以通过基于氧化物的记忆电阻器 Memristor（简称忆阻器）、自旋电子存储器、阈值开关和晶体管来实现。[5][7]对基于软件的脉冲神经网络系统的训练可以通过误差反向传播机制来实现，例如，使用snnTorch等基于Python的框架，[8]或使用BindsNet等典型的受生物启发的学习模式。[9]<br />
<br />
神经形态工程领域的一个关键问题，就是理解单个神经元形态、神经回路、应用和整体结构如何产生理想的计算，如何影响信息的表示和对破坏的鲁棒性，如何整合学习和发展，如何适应局部变化(可塑性) 并促进逐渐发展的变化。<br />
<br />
神经形态工程是以生物学、物理学、数学、计算机科学和电子工程[5]等学科为基础，设计人工神经系统（如视觉系统、头眼系统、听觉处理器以及物理结构和设计原则都受启发于生物神经系统的自主机器人）的一门交叉学科。[10]20世纪80年代后期，Carver Mead极大地推动了神经形态工程领域的发展。[11]<br />
<br />
<br />
== 发展脉络==<br />
<br />
===来源与启发===<br />
<br />
计算机已成为现代生活各个方面的关键。现如今，数据密集型应用对硬件性能提出了很高的要求，例如短访问延迟、大容量、高带宽、低成本和执行人工智能任务的能力。然而，来自日益增长的大数据压力带来了额外的挑战：一方面，由于复杂算法和架构的快速发展，功耗已成为一个显著的挑战。目前，世界上约5%-15%的能源用于各种形式的数据处理，如传输或加工任务，[54]并且考虑到物联网时代传感器无处不在，它们所生成的数据量也呈指数级增长，这一比例预计将迅速增加。另一方面，由于冯·诺依曼架构中处理单元和内存单元之间存在物理分离，基于此架构的数据处理越来越受到内存带宽的限制。虽然冯·诺依曼架构几十年来对科学技术的发展做出了巨大的贡献，但是由于相对缓慢且需要能量的数据移动，其表现出的效率是比较低的。<br />
<br />
基于互补金属氧化物半导体（CMOS）技术的传统冯·诺依曼计算机不具备像人脑那样的学习或处理复杂数据的能力。为了解决数字计算机的局限性，世界各地都在利用生物学原理的启发，开发截然不同的方法。其中一种方法就是模拟人脑中信息处理类型的计算系统，进行神经形态系统的开发。<br />
<br />
“神经形态”一词最初是由卡弗·米德在20世纪90年代创造的，指的是从大脑的神经生物结构中获得灵感的混合信号模拟/数字超大规模集成计算系统。[55] “神经形态工程”是一个跨学科研究领域，专注于构建电子神经处理系统，以直接“模拟”真实神经元和突触的生物物理特性。[56]<br />
<br />
神经元及其突触的生理学过程极其复杂，难以进行人工模拟。神经元在所有处理过程中都使用化学模拟信号，这是大脑的一个关键生理学特征。这个特征大大增加了在计算机中复制大脑的难度，因为目前的计算机是完全数字化的。然而，部分特征可以抽象为数学函数，这些函数能够紧密捕捉神经元运行的本质。<br />
<br />
最近，神经形态一词的定义在另外两个方向上进行了扩展。[57]首先，术语“神经形态”用于描述用于探索大规模计算神经科学模型的基于脉冲（spike）的处理系统。其次，神经形态计算包括实现神经元和突触电路的专用电子神经架构。注意，这一概念不同于人工智能机器学习方法，人工智能机器学习方法基于纯软件算法，开发纯软件算法是为了最小化模式识别任务中的识别错误。[58]然而，关于神经形态计算的精确定义有一些争议。它可以从非常严格的高保真模仿神经科学原理（其中必须有非常详细的突触化学动力学）到非常模糊的高层松散的大脑启发原理，例如简单的向量（输入）乘以矩阵（突触）乘法。<br />
<br />
总的来说，到今天为止，人们普遍认为神经形态计算至少应该包括一些时间、事件或数据驱动的计算。在这个意义上，像脉冲神经网络（SNN）这样的系统，有时被称为第三代神经网络，具有很强的代表性。[59]<br />
<br />
自然是向更可持续计算场景发展的重要灵感来源。在这种场景中，由于非易失性存储器、模拟/数字处理电路的集成等因素，神经形态系统具备复杂数据环境中的动态学习能力，显示出比传统处理器低得多的功耗。构建模拟生物对应物的人工神经网络是计算领域剩下的挑战之一。如果基本技术问题在未来几年内得到解决，低功耗、高速的神经形态计算机将推动对神经形态设备的需求，神经形态计算市场预计将从2025年的2亿美元上升到2035年的220亿美元。[60]<br />
<br />
神经形态计算的目标不是完美地模拟大脑及其所有功能，而是利用已知的大脑结构和运转机制来研发或优化实际的计算系统。任何神经形态学系统都不会声称或试图复制神经元和突触中的每一个元素，但所有人都一致认可将计算高度分散于一系列类似于神经元的小型计算元素的理念。研究人员用不同的方法来追求这一普遍目标。[12]<br />
<br />
<br />
===代表性成果===<br />
<br />
早在2006年，佐治亚理工学院的研究人员就研发出了一种现场可编程神经阵列。[13]在此之后，一系列越来越复杂的浮栅晶体管阵列被成功研发出来，这些晶体管阵列可以通过在金属-氧化物半导体效应晶体管MOSFET的栅极上编程来模拟大脑中神经元的离子通道特性，这也是以硅为主要材料的可编程神经元阵列的最早成功案例之一。<br />
<br />
2011年11月，麻省理工学院的一组研究人员研发出一种计算机芯片，该芯片上使用标准的互补金属氧化物半导体CMOS制造技术集成了400个晶体管，能够模拟神经元间突触中基于离子的通讯。[14][15]<br />
<br />
2012年6月，普渡大学的自旋电子学研究人员发表了一篇关于利用侧向自旋阀和忆阻器设计神经形态芯片的论文。他们认为，这种芯片结构的工作原理与神经元相似，因此可以用于大脑运行机制的复刻方法的测试。此外，这些芯片在能耗方面明显优于传统芯片。[16]<br />
<br />
惠普实验室在莫特忆阻器上的研究表明，尽管它们可以是非易失性的，但是在相变温度以下时表现出的易失性行为可以被用来制造类神经元电阻器（一种生物学启发的模仿神经元行为的硬件）。[17]2013年9月，他们通过模型和仿真展示了这些类神经元电阻器的脉冲行为如何产生图灵机的所需元素。[18]<br />
<br />
神经栅格 Neurogrid是由斯坦福大学Brains in Silicon公司研发的、使用神经形态工程原理设计的硬件。该电路板由16个定制设计的芯片组成。在设计中，每个NeuroCore芯片的模拟电路对65536个神经元的神经元素进行模拟，以最大限度地提高能量效率。模拟出的神经元通过设计的数字电路连接，以最大化脉冲吞吐量。[20][21]<br />
<br />
人类大脑计划 Human Brain Project对神经形态工程具有较大影响，其主要任务是尝试用生物数据在超级计算机中模拟完整的人脑。人类大脑计划由神经科学、医学和计算机科学背景的研究人员组成。[22]该项目的联合主管亨利•马克拉姆 Henry Markram表示，人类大脑计划的目的是建立一个探索和了解脑科学和脑疾病知识的基础，并利用这些知识来构建更先进的计算机技术。这个项目的三个主要目标分别是: 更好地理解大脑的各个部分是如何相互配合协同工作的; 理解如何客观地诊断和治疗脑部疾病; 以及利用对人类大脑的理解来开发神经形态计算机。模拟一个完整的人类大脑需要一台比现在强大一千倍的超级计算机，这不断激发着对神经形态计算机领域的研究兴趣。[23]欧盟委员会已经向人类大脑计划拨款13亿美元。[24]<br />
<br />
其他与神经形态工程有关的研究还包括脑计划 BRAIN initiative，[25]和IBM研发的TrueNorth芯片。[26]使用纳米晶体、纳米线和导电聚合物也能够用于制造神经形态学硬件。[27]<br />
<br />
2017年10月，英特尔发布了神经形态芯片 Loihi。该芯片采用异步脉冲神经网络实现了自适应、自修改、事件驱动的细粒度并行计算，实现了高效的学习和推理。[28][29]<br />
<br />
比利时的微电子研究中心（IMEC）研发了世界上首个自学习神经形态芯片。这种基于丝状氧化物电阻存储技术 filamentary-oxide-based resistive memory technology（OxRAM）技术的大脑启发芯片具有自学习能力，并且已被证明具有创作音乐的能力。[30]IMEC发布了由芯片原型机谱写的30秒旋律。向芯片加载一系列特征、风格统一的歌曲（古代比利时和法国长笛小步舞曲），芯片从中学习相关规则并将其应用于创作。[31]<br />
<br />
由Henry Markram领导的蓝脑计划 The Blue Brain Project旨在建立小鼠大脑生理学细节的数字重建和模拟。蓝脑计划已经建立了啮齿动物大脑的电子模型，同时进行着尽可能多地复制其生理学细节的尝试。基于超级计算机的模拟为理解大脑的结构和功能提供了新的视角。<br />
<br />
欧盟资助了海德堡大学一系列促进BrainScaleS（神经形态混合系统中受大脑启发的多尺度计算）发展的项目，这是一台位于德国海德堡大学的混合模拟神经形态超级计算机。它是作为人类大脑计划中神经形态计算平台的一部分而开发的，是SpiNNaker超级计算机（基于数字技术）的补充。BrainScaleS中使用的体系架构模拟了生物神经元及其在物理层面上的连接；此外，由于这些组件是由硅制成的，这些模型神经元平均运行速度是生物神经元的864倍，这意味着在机器模拟中，24小时的实时时间仅为100秒。[32]<br />
<br />
<br />
== 神经形态材料和设备 ==<br />
<br />
为了推进神经形态计算和工程领域，探索新型材料和器件将具有关键意义，以便以颠覆性的方式提高最先进的CMOS解决方案的能效和可扩展性。[57][61]<br />
<br />
忆阻器件可以改变其电导以响应电脉冲，[62][63][64]有望作为神经形态电路中突触和神经元的能量和空间高效硬件表征。忆阻器件最初是作为二进制非易失性随机存取存储器提出的，该领域的研究目标主要是在固态驱动器技术（例如闪存更换）或存储类存储器[65]中寻求更高性能。然而，由于模拟可调性和复杂的开关动力学，忆阻器件还可以实现新的计算功能，例如模拟计算或实现受大脑启发的学习规则。据报道，大量不同的物理现象都表现出忆阻行为，包括电子效应、离子效应以及结构有序效应。<br />
<br />
神经形态工程中使用的材料包括磁性合金、金属氧化物、硫属化合物，也包括2D 范德瓦尔斯材料或有机材料。部分新兴存储设备已经作为二进制非易失性存储器有了商业应用，例如相变存储器（PCM）、磁性随机存取存储器、铁电存储器以及基于氧化还原的电阻随机存取存储器。纳米线网络（nanowire networks）、2D材料、有机材料等材料的发展尚未成熟，但可能为柔性电子或3D集成提供扩展功能和新机会。<br />
<br />
<br />
===Intel芯片Loihi及系统Lava[49][50][51]===<br />
<br />
Loihi是Intel实验室设计的一种神经形态研究测试芯片，它使用异步脉冲神经网络（SNN）来实现自适应、自修改、事件驱动的细粒度并行计算，用于高效实现学习和推理。该芯片是一个128个神经形态核的多核集成电路，采用英特尔14纳米工艺制造，具有独特的用于片上SNN训练的可编程微码学习引擎。该芯片在俄勒冈州2018年神经启发计算元素（NICE）研讨会上正式亮相。<br />
<br />
2021年9月，Intel推出其第二代神经形态研究芯片Loihi2和用于神经形态应用研发的开源软件框架Lava。<br />
<br />
Loihi 2的进步使该架构能够支持新类别的神经形态算法和应用，同时提供高达10倍的处理速度，高达15倍的资源密度。每个芯片上有多达100万个神经元，并提高了能效。与过去的工艺技术相比，Loihi2采取Intel 4技术，使用极紫外光刻技术，简化了布局设计规则，使得Loihi 2的快速开发成为可能。<br />
<br />
Lava软件框架解决了神经形态研究社区对通用软件框架的需求。作为一个开放、模块化和可扩展的框架，Lava将允许研究人员和应用程序开发人员在彼此的基础上进行开发，并提供一组通用的工具、方法和库来进行聚合。Lava可在传统处理器和神经形态处理器的异构架构上无缝运行，实现了各种人工智能、神经形态和机器人框架的跨平台执行和互操作。开发人员可以直接开始构建神经形态应用程序，而无需访问专门的神经形态硬件，并可以为Lava代码库做出贡献，包括将其移植到其他平台上运行。<br />
<br />
<br />
===清华天机芯Tianjic[52]===<br />
<br />
天机芯片由清华类脑计算中心团队推出，结合了基于计算机科学和基于神经科学的两种人工通用智能开发方法，提供了一个混合、协同的平台。Tianjic芯片采用了多核架构、可重构的构建块和带有混合编码方案的简化数据流，不仅可以适应基于计算机科学的机器学习算法，还可以轻松实现脑激励电路和多种编码方案。研发团队已经实现在实时目标检测、跟踪、语音控制、避障和平衡控制等任务中仅使用该芯片同时处理多种算法和模型的尝试。天机芯片的进一步研究有望通过促进通用硬件平台的发展来促进通用人工智能的发展。<br />
<br />
<br />
===神经忆阻系统===<br />
<br />
神经忆阻系统是神经形态计算系统的一个亚类，主要研究利用忆阻器 Memristors实现神经可塑性 Neuroplasticity。神经形态工程的重点是模拟生物行为，而神经忆阻系统的重点是提取。<ref name=":33">{{Cite web|url=https://digitalops.sandia.gov/Mediasite/Play/a10cf6ceb55d47608bb8326dd00e46611d|title=002.08 N.I.C.E. Workshop 2014: Towards Intelligent Computing with Neuromemristive Circuits and Systems - Feb. 2014|website=digitalops.sandia.gov|access-date=2019-08-26}}</ref>举个例子，一个神经忆阻系统可能用抽象的神经网络模型替代皮层微电路的行为细节。<ref name=":34">C. Merkel and D. Kudithipudi, "Neuromemristive extreme learning machines for pattern classification," ISVLSI, 2014.</ref><br />
<br />
<br />
受神经元启发、使用忆阻器实现的阈值逻辑函数<ref name="Maan 1–13" />在高级模式识别中有着广泛的应用，最近报道中其应用包括语音识别<ref name=":35">{{Cite journal|title = Memristor pattern recogniser: isolated speech word recognition|journal = Electronics Letters|pages = 1370–1372|volume = 51|issue = 17|doi = 10.1049/el.2015.1428|first1 = A.K.|last1 = Maan|first2 = A.P.|last2 = James|first3 = S.|last3 = Dimitrijev|year = 2015|bibcode = 2015ElL....51.1370M|hdl = 10072/140989|url = https://semanticscholar.org/paper/48d3ab11ec6e213b62f11eedcfb7b7febb058674|hdl-access = free}}</ref>、人脸识别<ref name=":36">{{Cite journal|title = Memristive Threshold Logic Face Recognition|journal = Procedia Computer Science|date = 2014-01-01|pages = 98–103|volume = 41|series = 5th Annual International Conference on Biologically Inspired Cognitive Architectures, 2014 BICA|doi = 10.1016/j.procs.2014.11.090|first1 = Akshay Kumar|last1 = Maan|first2 = Dinesh S.|last2 = Kumar|first3 = Alex Pappachen|last3 = James|doi-access = free}}</ref>和物体识别<ref name=":37">{{Cite journal|title = Memristive Threshold Logic Circuit Design of Fast Moving Object Detection|journal = IEEE Transactions on Very Large Scale Integration (VLSI) Systems|date = 2015-10-01|issn = 1063-8210|pages = 2337–2341|volume = 23|issue = 10|doi = 10.1109/TVLSI.2014.2359801|first1 = A.K.|last1 = Maan|first2 = D.S.|last2 = Kumar|first3 = S.|last3 = Sugathan|first4 = A.P.|last4 = James|arxiv = 1410.1267}}</ref>。阈值逻辑函数还可以用来取代传统的数字逻辑门。<ref name=":38">{{Cite journal|title = Resistive Threshold Logic|journal = IEEE Transactions on Very Large Scale Integration (VLSI) Systems|date = 2014-01-01|issn = 1063-8210|pages = 190–195|volume = 22|issue = 1|doi = 10.1109/TVLSI.2012.2232946|first1 = A.P.|last1 = James|first2 = L.R.V.J.|last2 = Francis|first3 = D.S.|last3 = Kumar|arxiv = 1308.0090}}</ref><ref name=":39">{{Cite journal|title = Threshold Logic Computing: Memristive-CMOS Circuits for Fast Fourier Transform and Vedic Multiplication|journal = IEEE Transactions on Very Large Scale Integration (VLSI) Systems|date = 2015-11-01|issn = 1063-8210|pages = 2690–2694|volume = 23|issue = 11|doi = 10.1109/TVLSI.2014.2371857|first1 = A.P.|last1 = James|first2 = D.S.|last2 = Kumar|first3 = A.|last3 = Ajayan|arxiv = 1411.5255}}</ref><br />
<br />
<br />
对于理想的无源记忆电路，电路的内部记忆可以用精确的方程（Caravelli-Traversa-Di Ventra方程) 来描述:<ref name=":40">{{cite journal |last=Caravelli |display-authors=etal|arxiv=1608.08651 |title=The complex dynamics of memristive circuits: analytical results and universal slow relaxation |year=2017 |doi=10.1103/PhysRevE.95.022140 |pmid= 28297937 |volume=95 |issue= 2 |pages= 022140 |journal=Physical Review E|bibcode=2017PhRvE..95b2140C}}</ref><br />
<br />
<br />
:<math> \frac{d}{dt} \vec{W} = \alpha \vec{W}-\frac{1}{\beta} (I+\xi \Omega W)^{-1} \Omega \vec S </math><br />
<br />
<br />
Caravelli-Traversa-Di Ventra方程是描述物理记忆网络和外部源性质的函数。在上述方程中，<math>\alpha</math>是“遗忘”时间尺度常数，<math>\xi=r-1</math>，<math>r =\frac{R\text_{off}}{R_\text{on}}</math>是记忆电阻器off状态和on状态极限电阻值之比，<math>\vec S</math>是电路源的矢量，<math>\Omega</math>是电路基本环路的投影。常数<math>\beta</math>具有电压的量纲，与记忆电阻器的特性有关；其物理原型是导体中的电荷迁移率。对角矩阵和向量 <math>W=\operatorname{diag}(\vec W)</math>和<math>\vec W</math> 分别是忆阻器的内阻，值在0到1之间。因此，这个等式需要在内存值=上添加额外约束以保证可靠性。<br />
<br />
<br />
==神经形态电路==<br />
<br />
神经形态器件可以与传统的CMOS晶体管集成，以开发功能齐全的神经形态电路。神经形态电路中的一个关键元素是非冯·诺依曼架构，例如由多个核组成，实现分布式的计算和内存。采用脉冲来表示、交换和计算类似于大脑动作电位的数据的SSN，和受大脑启发程度较小的电路（例如ANN），通常都包括在神经形态电路中。<br />
<br />
无论具体学习和处理算法是什么，神经形态电路中的一个关键处理元素是神经网络，包括突触和神经元。鉴于神经网络的核心作用，目前一项重要的研究工作就是如何通过内存计算（inmemory computing）实现密集、快速和节能的神经网络。[66]例如，内存阵列可以加速矩阵向量乘法（MVM），[67]并利用欧姆定律和基尔霍夫定律在网络中实现乘法和求和，这是许多神经形态电路的共同特征，包括SNN和非SNN网络。MVM交叉点电路能够以高密度、高实时处理速度和高能效完成对突触层的直接实现，尽管其准确性受到记忆器件随机变化和模拟计算的挑战。<br />
<br />
另一个电路挑战是模数混合计算，这导致需要在模拟交叉点阵列和数字系统之间的接口处需要使用大型高能模数转换器电路。<br />
<br />
最后，神经形态电路似乎从混合集成中能取得最大的获益，即将前端CMOS技术与能够实现MVM和神经生物学功能的新型存储设备相结合，例如脉冲集成、短期记忆和突触可塑性。[68]从长远来看，混合集成可能还需要扩展到替代性纳米技术概念，例如自下而上的纳米线网络，[69]以及替代性计算概念，例如光子[70]甚至量子计算[71]，在单个系统甚至单个芯片内具有三维集成结构。<br />
<br />
<br />
==神经形态算法==<br />
<br />
在真实应用系统应用神经形态工程的一个基本挑战是要直接在脉冲域中对其进行训练，以便达到更节能、更精确的效果，实现在不依赖重型云计算服务器的情况下不断学习和更新便携式设备本身的知识也是一个挑战。脉冲数据往往分布稀疏，具有一定的随机性和嵌入噪声，与非理想非线性突触和神经元相互作用。生物知道如何充分利用这一切，并从周围环境中有效地获取知识。<br />
<br />
从这个意义上说，计算神经科学可以成为启发神经形态工程学的关键因素：这门学科中大脑如何在各种尺度上执行计算，从小神经元群、中尺度聚合到完整组织，大脑区域和整个大脑与周围传感器和运动执行器相互作用，都是非常值得学习的。另一方面，关于信息如何在大脑中使用神经脉冲编码的基本问题也出现了。显然，为了提高处理和通信的能效，大脑会最大限度地提高每单位脉冲的信息量。[72]这意味着通过利用时空信号处理来分解信息的编码和处理，以最大化信息量，同时最小化能量和资源。<br />
<br />
<br />
===SNN===<br />
<br />
受到生物神经元特性启发，脉冲神经网络（SNN）使用脉冲——这是一种发生在时间点上的离散事件——而非常见的连续值。每个峰值由代表生物过程的微分方程表示出来，其中最重要的是神经元的膜电位。本质上，一旦神经元达到了某一电位，脉冲就会出现，随后达到电位的神经元会被重置。SNN电路的动态特性及其由异步技术支持的时空稀疏性使其在快速有效地处理动态信号方面特别有前途。<br />
<br />
与成熟有效的人工神经网络 (ANN) 训练算法：误差反向传播算法 (Back Propagation) 不同，复杂的动力学和脉冲的不可微性质导致SNN的训练更为困难。此外，神经形态脉冲神经结构无法如冯·诺伊曼架构般访问外部存储块来存储或检索信息。<br />
<br />
在理解如何在硬件SNN中最佳设计信号处理和在线学习机制以解决复杂问题方面，神经形态计算和工程领域面临的挑战似乎是无法克服的。然而，生物大脑是一种存在证明，可以使用模拟的计算基底，并使用不均匀和不精确的信号处理元件，实现稳健和稳定的计算。因此，了解动物的大脑，即使是只有不到一百万个神经元的小昆虫大脑，是如何完成这项任务的，这也是在这一领域取得进展的关键。具体来说，应对这些挑战所需的科学进步需要强有力的跨学科方法。理论和计算神经科学的进步将提供一个核心组成部分。但是，这些需要得到电气工程多个子领域的概念和结果的补充，如信息理论、信号处理和控制理论，以及其他学科，如数学、计算机科学和机器人学。<br />
<br />
<br />
==应用==<br />
<br />
尽管神经形态计算和工程仍处于研发阶段，但随着研究进展，越来越多的应用领域被证明非常适合神经形态计算和工程，确定新的潜在应用领域以进一步探索使用神经形态材料和硬件如何解决这些问题也很重要。<br />
<br />
===视觉===<br />
<br />
基于神经形态事件（event-based, EB）的视觉传感器从人类视网膜的结构与功能中获得灵感，试图在超大规模集成电路硅芯片上重建视网膜的视觉信息采集和处理操作。加州理工学院C·米德团队研发的第一台此类设备名为“硅视网膜”，它于1991年登上了《科学美国人》的封面。[73]与早期更可靠的生物学模型相比，近年来，人们更多地关注创建可用于实际人工视觉应用的实际传感器设计，这些模型通常对许多不同的细胞类型和信号通路进行建模，从而导致设计方案非常复杂且实际可用性有限。<br />
<br />
如今，大多数EB传感器设备基于“时间对比”或“变化检测”（change detection, CD）类型的操作，松散地模拟人类视觉系统的瞬态磁细胞路径。与传统图像传感器不同，CD传感器不对所有像素使用一个公共采样率（帧速率），而是每个像素通过对入射光量的变化作出反应来定义其自身采样点的定时，以响应其视觉输入。[74][75][76]因此，整个采样过程不再由人工定时源控制，而是由待采样信号本身控制，或者更准确地说，由信号随时间的变化控制。这种传感器生成的输出不是图像序列，而是基于场景中发生的动态有条件地生成和传输的像素单个对比度事件的准时间连续流。<br />
<br />
由于所获得视觉动态的高时间精度、固有的数据稀疏性和鲁棒的高动态范围操作，加上在功率预算、后处理能力或传输带宽方面的考虑，EB传感器作为人工视觉系统的视觉传感器在需要高速或低延迟操作的应用、照明条件不受控制、资源有限的应用场景中中越来越普遍，例如在各种汽车、物联网、监控、移动或工业用例中。[77]<br />
<br />
<br />
===听觉===<br />
<br />
神经形态听觉技术的灵感来自人类的听觉能力，人类即使在困难的听觉场景中也能理解语音，并且只使用大脑整个10W功耗的一小部分。匹配人类听觉能力是开发人工听觉设备中算法、硬件技术和应用程序的一个重要目标。<br />
<br />
人类的听觉始于生物耳蜗，它使用空间来进行编码。传入的声音被编码为一组宽频率选择性通道产生的异步输出脉冲。对于频率低于3 kHz的信号，这些脉冲被锁相到频率[78]。这种编码方案导致从活动频率通道对频率信息进行稀疏采样，而不是在单个音频输入上使用最大采样率。最早的硅耳蜗设计始于Lyon和Mead（electronic cochlea）的工作，即通过一组耦合的过滤层级来模拟耳蜗的基底膜（BM）。后续设计中，滤波器层级具的设计有更好匹配特性，使用的耦合滤波器架构范围从最初提出的级联型对耳蜗的唯象输出的建模，[79]到更明确地模拟BM和耳蜗液作用的电阻耦合带通滤波器组。[80][81]最近推出的采用更先进技术的脉冲耳蜗设计取得了更好的功率效率。<br />
<br />
神经形态仿生特征，例如稀疏采样（例如，沉默暂停期间的非采样）、事件驱动的大脑计算形式、异步事件（指定时间窗口内的脉冲或脉冲集）携带的自然时间编码等，可以为解决听觉任务的硬件友好型模型提供更节能的解决方案。<br />
<br />
<br />
==社会与伦理==<br />
<br />
===伦理问题===<br />
<br />
虽然神经形态工程这个跨学科概念相对较新，但许多适用于类人机器 Human-like machines和人工智能的伦理讨论在神经形态系统领域也无法避免。另外，神经形态系统是为了模仿人类大脑而设计，这一底层逻辑也导致了一些新的伦理问题。<br />
<br />
然而，这种争论的实际情况是，神经形态硬件和人工“神经网络”是对大脑运作或处理信息过程的极其简化的模型，在大小和功能技术方面的复杂性相比而言比较低，在连接性方面也具有更加规则的结构。将神经形态芯片与大脑进行比较是一种非常粗糙的比较，类似于仅仅因为飞机有翅膀和尾巴就将它与鸟进行比较。事实上，当前最先进的人工智能在能耗效率和计算效率方面距离人脑神经认知系统仍有较大差距，而神经形态工程只是一种通过从大脑机制中获得灵感来缩小这种差距的尝试，就像许多工程设计中都具有生物启发的特征 Bio-inspired features一样。<br />
<br />
由于公众认知的相关忧虑，神经形态工程学可能会受到严重的伦理限制。[33]欧盟委员会进行的一项调查指出，公众越来越关注能够模仿或复制人类行为的机器人。[34]而神经形态工程，顾名思义，是为了模仿人脑的功能而设计的。因此，神经形态系统可能因为其类似人类的特性而被归入许多欧盟公民希望在未来禁止使用的机器人类别。<br />
<br />
<br />
====人格权问题====<br />
<br />
随着神经形态系统的日益发展，一些学者主张赋予这些系统人格权。如果是大脑赋予了人类人格，那么在多大程度上模仿人类大脑的神经形态系统才能被赋予人格权利？“人类大脑计划”旨在推进以大脑为灵感的计算机技术发展，该计划的批评者认为，神经形态计算机技术的进步可能导致机器意识或人格的形成。[35]这些批评者认为，如果这些系统被当作人来对待，那么人类使用神经形态系统执行任务（包括终止神经形态系统）的行为，在道德上就可能是不被允许的，因为这些行为将违反神经形态系统的自主性。[36]<br />
<br />
===法律问题===<br />
<br />
怀疑派认为，在法律上没有办法应用能够适用于神经形态技术的电子人格。在一封由285名法律、机器人技术、医学和伦理学专家的联名信中，作者们反对欧盟委员会提出的承认“智能机器人”为法人的提议。他们写道，“机器人的法律地位无法从自然人模型中推导出来，因为机器人将被赋予人权，如尊严权、完整权、报酬权或公民权，从而直接面临人权问题。这将有悖于《欧联基本权利宪章》和《欧洲保障人权和根本自由公约》”。[37]<br />
<br />
====所有权及财产权问题====<br />
<br />
法律界围绕财产权和人工智能有着重大争论。在Acohs Pty有限公司诉Ucorp Pty有限公司一案中，澳大利亚联邦法院的Christopher Jessup法官发现，版权保护不适用于材料安全数据表的源代码，因为它是由软件界面生成而非人类工作者生成的。[38]同样的问题可能也适用于神经形态系统：如果一个神经形态系统成功地模仿了人类的大脑，并产生了一部原创作品，那么该如何确认这部作品的所有权归属？[39]<br />
<br />
===军民两用技术===<br />
<br />
联合人工智能中心 The Joint Artificial Intelligence Center（JAIC），是美国军队的一个分支，专门从事采购和实施用于战斗的人工智能软件和神经形态硬件。具体应用包括智能耳机、护目镜和机器人。JAIC打算高度依赖神经形态技术，使用神经形态技术来连接神经形态单位网络中的“每个战士、每个射手”。<br />
<br />
<br />
'''神经形态工程 Neuromorphic engineering'''（也称为'''神经形态计算 Neuromorphic computing'''或'''类脑计算 Brain-inspired computing'''）<ref name=":3">{{Cite journal | doi = 10.1145/2601069| title = Neuromorphic computing gets ready for the (really) big time| journal = [[Communications of the ACM]]| volume = 57| issue = 6| year = 2014| last1 = Monroe | first1 = D. | pages = 13–15}}</ref><ref name=":4">{{Cite journal | doi = 10.1088/0957-4484/21/17/175202| pmid = 20368686| title = Nanotube devices based crossbar architecture: Toward neuromorphic computing| journal = Nanotechnology| volume = 21| issue = 17| pages = 175202| year = 2010| last1 = Zhao | first1 = W. S. | last2 = Agnus | first2 = G. | last3 = Derycke | first3 = V. | last4 = Filoramo | first4 = A. | last5 = Bourgoin | first5 = J. -P. | last6 = Gamrat | first6 = C. | bibcode = 2010Nanot..21q5202Z| url = https://zenodo.org/record/3428659}}</ref><ref name="humanbrainproject">{{YouTube|id=6RoiZ90mGfw|title=The Human Brain Project SP 9: Neuromorphic Computing Platform}}</ref>是指使用包含电子模拟电路的超大规模集成电路系统来模拟神经系统中生理结构的研究方法。神经形态计算机或神经形态芯片包括任何使用由硅制成的人造神经元进行计算的设备。<ref name=":5">{{cite journal|last1=Mead|first1=Carver|title=Neuromorphic electronic systems|journal=Proceedings of the IEEE|date=1990|volume=78|issue=10|pages=1629–1636|doi=10.1109/5.58356|url=https://authors.library.caltech.edu/53090/1/00058356.pdf}}</ref><ref name=":2" />近年来，'''神经形态学 neuromorphic'''被用来描述能够实现神经系统模型功能（如感知、运动控制，多感官整合等）的模拟、数字、模拟/数字混合模式超大规模集成电路和软件系统。神经形态计算的硬件实现可以通过基于氧化物的记忆电阻器 Memristor（简称忆阻器）、<ref name="Maan 1–13">{{Cite journal|last1=Maan|first1=A. K.|last2=Jayadevi|first2=D. A.|last3=James|first3=A. P.|date=2016-01-01|title=A Survey of Memristive Threshold Logic Circuits|journal=IEEE Transactions on Neural Networks and Learning Systems|volume=PP|issue=99|pages=1734–1746|doi=10.1109/TNNLS.2016.2547842|pmid=27164608|issn=2162-237X|arxiv=1604.07121|bibcode=2016arXiv160407121M}}</ref>自旋电子存储器、阈值开关和晶体管来实现。<ref name=":6">{{Cite journal|title = Mott Memory and Neuromorphic Devices|journal = Proceedings of the IEEE|date = 2015-08-01|issn = 0018-9219|pages = 1289–1310|volume = 103|issue = 8|doi = 10.1109/JPROC.2015.2431914|first1 = You|last1 = Zhou|first2 = S.|last2 = Ramanathan|url = https://zenodo.org/record/895565}}</ref><ref name=":2">{{Cite journal|title=Neuromorphic Circuits With Neural Modulation Enhancing the Information Content of Neural Signaling {{!}} International Conference on Neuromorphic Systems 2020|language=EN|doi=10.1145/3407197.3407204}}</ref>对基于软件的脉冲神经网络系统的训练可以通过误差反向传播机制来实现，例如，使用snnTorch等基于Python的框架，<ref name=":7">{{cite journal |last1=Eshraghian|first1=Jason K.|last2=Ward|first2=Max|last3=Neftci |first3=Emre|last4=Wang|first4=Xinxin|last5=Lenz|first5=Gregor|last6=Dwivedi|first6=Girish|last7=Bennamoun|first7=Mohammed|last8=Jeong|first8=Doo Seok|last9=Lu|first9=Wei D.|title=Training Spiking Neural Networks Using Lessons from Deep Learning |date=1 October 2021 |arxiv=2109.12894 }}</ref> 或使用BindsNet等典型的受生物启发的学习模式。<ref name=":8">{{Cite web | url=https://github.com/Hananel-Hazan/bindsnet | title=Hananel-Hazan/bindsnet: Simulation of spiking neural networks (SNNs) using PyTorch.| date=31 March 2020}}</ref><br />
<br />
<br />
神经形态工程领域的一个关键问题，就是理解单个神经元形态、神经回路、应用和整体结构如何产生理想的计算，如何影响信息的表示和对破坏的鲁棒性，如何整合学习和发展，如何适应局部变化(可塑性) 并促进逐渐发展的变化。<br />
<br />
<br />
神经形态工程是以生物学、物理学、数学、计算机科学和电子工程<ref name=":2" />等学科为基础，设计人工神经系统（如视觉系统、头眼系统、听觉处理器以及物理结构和设计原则都受启发于生物神经系统的自主机器人）的一门交叉学科。<ref name=":9">{{Cite journal | doi = 10.1155/2012/705483| title = Qualitative Functional Decomposition Analysis of Evolved Neuromorphic Flight Controllers| journal = Applied Computational Intelligence and Soft Computing| volume = 2012| pages = 1–21| year = 2012| last1 = Boddhu | first1 = S. K. | last2 = Gallagher | first2 = J. C. | doi-access = free}}</ref>20世纪80年代后期，Carver Mead极大地推动了神经形态工程领域的发展。<ref name=":10">{{cite web|title=carver mead website|last1=Mead|first1=Carver|url=http://www.carvermead.caltech.edu/index.html|website=carvermead}}</ref><br />
<br />
<br />
==神经生物学启发==<br />
神经形态工程形成的灵感来源于现已知的大脑结构及运作机制知识，它将我们对大脑功能的了解用于研究和优化计算机系统。该领域工作主要集中于对生物计算模拟特性和神经元在认知中所发挥作用的复刻。<br />
<br />
<br />
神经元及其[[突触]]的生理学过程极其复杂，难以进行人工模拟。神经元在所有处理过程中都使用化学模拟信号，这是大脑的一个关键生理学特征。这个特征大大增加了在计算机中复制大脑的难度，因为目前的计算机是完全数字化的。然而，部分特征可以抽象为数学函数，这些函数能够紧密捕捉神经元运行的本质。<br />
<br />
<br />
神经形态计算的目标不是完美地模拟大脑及其所有功能，而是利用已知的大脑结构和运转机制来研发或优化实际的计算系统。任何神经形态学系统都不会声称或试图复制神经元和突触中的每一个元素，但所有人都一致认可将计算高度分散于一系列类似于神经元的小型计算元素的理念。研究人员用不同的方法来追求这一普遍目标。<ref name=":11">{{Cite journal | doi = 10.1088/1741-2560/13/5/051001| title = Large-scale neuromorphic computing systems| journal = Journal of Neural Engineering| volume = 13| pages = 1–15| year = 2016| last1 = Furber | first1 = Steve| issue = 5| pmid = 27529195| bibcode = 2016JNEng..13e1001F| doi-access = free}}</ref><br />
<br />
<br />
==代表性成果==<br />
早在2006年，佐治亚理工学院的研究人员就研发出了一种现场可编程神经阵列。<ref name=":12">{{Cite book|title = A field programmable neural array|last1 = Farquhar|first1 = Ethan|date = May 2006|journal = IEEE International Symposium on Circuits and Systems|pages = 4114–4117|last2 = Hasler|first2 = Paul.|doi = 10.1109/ISCAS.2006.1693534|isbn = 978-0-7803-9389-9}}</ref>在此之后，一系列越来越复杂的浮栅晶体管阵列被成功研发出来，这些晶体管阵列可以通过在'''金属-氧化物半导体效应晶体管MOSFET'''的栅极上编程来模拟大脑中神经元的离子通道特性，这也是以硅为主要材料的可编程神经元阵列的最早成功案例之一。<br />
<br />
<br />
2011年11月，麻省理工学院的一组研究人员研发出一种计算机芯片，该芯片上使用标准的'''互补金属氧化物半导体CMOS'''制造技术集成了400个晶体管，能够模拟神经元间突触中基于离子的通讯。<ref name=":13">{{cite web|title=MIT creates "brain chip"|url=http://www.extremetech.com/extreme/105067-mit-creates-brain-chip|access-date=4 December 2012}}</ref><ref name="Neuromorphic silicon paper">{{cite journal|title=Neuromorphic silicon neurons and large-scale neural networks: challenges and opportunities|doi=10.3389/fnins.2011.00108|pmid=21991244|pmc=3181466|volume=5|pages=108|journal=Frontiers in Neuroscience|year=2011|last1=Poon|first1=Chi-Sang|last2=Zhou|first2=Kuan|doi-access=free}}</ref><br />
<br />
<br />
2012年6月，普渡大学的自旋电子学研究人员发表了一篇关于利用侧向自旋阀和忆阻器设计神经形态芯片的论文。他们认为，这种芯片结构的工作原理与神经元相似，因此可以用于大脑运行机制的复刻方法的测试。此外，这些芯片在能耗方面明显优于传统芯片。<ref name="Spin Devices Prop">{{Cite arXiv|title=Proposal For Neuromorphic Hardware Using Spin Devices|eprint=1206.3227|last1=Sharad|first1=Mrigank|last2=Augustine|first2=Charles|last3=Panagopoulos|first3=Georgios|last4=Roy|first4=Kaushik|class=cond-mat.dis-nn|year=2012}}</ref><br />
<br />
<br />
惠普实验室在莫特忆阻器上的研究表明，尽管它们可以是非易失性的，但是在[[相变]]温度以下时表现出的易失性行为可以被用来制造类神经元电阻器（一种生物学启发的模仿神经元行为的硬件）<ref name=":0">{{Cite journal | doi = 10.1038/nmat3510| pmid = 23241533| title = A scalable neuristor built with Mott memristors| journal = Nature Materials| volume = 12| issue = 2| pages = 114–7| year = 2012| last1 = Pickett | first1 = M. D. | last2 = Medeiros-Ribeiro | first2 = G. | last3 = Williams | first3 = R. S. | bibcode = 2013NatMa..12..114P| url = https://semanticscholar.org/paper/b6ba6f496ace2b947f111059663e76bb60e9efeb}}</ref>。2013年9月，他们通过模型和仿真展示了这些类神经元电阻器的脉冲行为如何产生[[图灵机]]的所需元素。<ref name=":14">{{cite journal|doi=10.1088/0957-4484/24/38/384002|title=Phase transitions enable computational universality in neuristor-based cellular automata|author1=Matthew D Pickett|author2=R Stanley Williams|date=September 2013|publisher=IOP Publishing Ltd|journal=Nanotechnology|volume=24|issue=38|pmid=23999059|bibcode=2013Nanot..24L4002P|at=384002}}</ref><br />
<br />
<br />
'''神经栅格 Neurogrid<ref name=":15">{{cite journal|last1=Boahen|first1=Kwabena|title=Neurogrid: A Mixed-Analog-Digital Multichip System for Large-Scale Neural Simulations|journal=Proceedings of the IEEE|date=24 April 2014|volume=102|issue=5|pages=699–716|doi=10.1109/JPROC.2014.2313565}}</ref>'''是由斯坦福大学Brains in Silicon公司研发的、使用神经形态工程原理设计的硬件。该电路板由16个定制设计的芯片组成。在设计中，每个NeuroCore芯片的模拟电路对65536个神经元的神经元素进行模拟，以最大限度地提高能量效率。模拟出的神经元通过设计的数字电路连接，以最大化脉冲吞吐量。<ref name=":16">{{cite journal|doi=10.1038/503022a|pmid = 24201264|title = Neuroelectronics: Smart connections|journal = Nature|volume = 503|issue = 7474|pages = 22–4|year = 2013|last1 = Waldrop|first1 = M. Mitchell|bibcode = 2013Natur.503...22W|doi-access = free}}</ref><ref name=":17">{{cite journal|doi=10.1109/JPROC.2014.2313565|title = Neurogrid: A Mixed-Analog-Digital Multichip System for Large-Scale Neural Simulations|journal = Proceedings of the IEEE|volume = 102|issue = 5|pages = 699–716|year = 2014|last1 = Benjamin|first1 = Ben Varkey|last2 = Peiran Gao|last3 = McQuinn|first3 = Emmett|last4 = Choudhary|first4 = Swadesh|last5 = Chandrasekaran|first5 = Anand R.|last6 = Bussat|first6 = Jean-Marie|last7 = Alvarez-Icaza|first7 = Rodrigo|last8 = Arthur|first8 = John V.|last9 = Merolla|first9 = Paul A.|last10 = Boahen|first10 = Kwabena}}</ref><br />
<br />
<br />
人类大脑计划 Human Brain Project对神经形态工程具有较大影响，其主要任务是尝试用生物数据在超级计算机中模拟完整的人脑。人类大脑计划由神经科学、医学和计算机科学背景的研究人员组成。<ref name=":18">{{cite web|title=Involved Organizations|url=http://www.humanbrainproject.eu/partners.html|access-date=22 February 2013|url-status=dead|archive-url=https://web.archive.org/web/20130302142627/http://www.humanbrainproject.eu/partners.html|archive-date=2 March 2013}}</ref>该项目的联合主管亨利•马克拉姆 Henry Markram表示，人类大脑计划的目的是建立一个探索和了解脑科学和脑疾病知识的基础，并利用这些知识来构建更先进的计算机技术。这个项目的三个主要目标分别是: 更好地理解大脑的各个部分是如何相互配合协同工作的; 理解如何客观地诊断和治疗脑部疾病; 以及利用对人类大脑的理解来开发神经形态计算机。模拟一个完整的人类大脑需要一台比现在强大一千倍的超级计算机，这不断激发着对神经形态计算机领域的研究兴趣。<ref name=":19">{{cite web|title=Human Brain Project|url=http://www.humanbrainproject.eu|access-date=22 February 2013}}</ref>欧盟委员会已经向人类大脑计划拨款13亿美元。<ref name=":20">{{cite web|title=The Human Brain Project and Recruiting More Cyberwarriors|url=http://www.marketplace.org/topics/tech/human-brain-project-and-recruiting-more-cyberwarriors|access-date=22 February 2013|date=January 29, 2013}}</ref><br />
<br />
<br />
<br />
其他与神经形态工程有关的研究还包括脑计划 BRAIN initiative，<ref name="economist">[https://www.economist.com/news/science-and-technology/21582495-computers-will-help-people-understand-brains-better-and-understanding-brains Neuromorphic computing: The machine of a new soul], The Economist, 2013-08-03</ref> 和IBM研发的TrueNorth芯片。<ref name=":21">{{cite journal|last1=Modha|first1=Dharmendra|title=A million spiking-neuron integrated circuit with a scalable communication network and interface|journal=Science|date=Aug 2014|volume=345|issue=6197|pages=668–673|doi=10.1126/science.1254642|pmid=25104385|bibcode=2014Sci...345..668M}}</ref>使用纳米晶体、纳米线和导电聚合物也能够用于制造神经形态学硬件。<ref name=":22">{{Cite web|url=http://jessamynfairfield.com/wp-content/uploads/2017/03/PWMar17Fairfield.pdf|title=Smarter Machines|last=Fairfield|first=Jessamyn|date=March 1, 2017}}</ref><br />
<br />
<br />
<br />
2017年10月，英特尔发布了神经形态芯片 Loihi。该芯片采用异步[[脉冲神经网络]]实现了自适应、自修改、事件驱动的细粒度并行计算，实现了高效的学习和推理。<ref name=":23">{{cite journal |last1=Davies |first1=Mike |title=Loihi: A Neuromorphic Manycore Processor with On-Chip Learning |journal=IEEE Micro |date=January 16, 2018 |volume=38 |issue=1 |pages=82–99 |display-authors=etal|doi=10.1109/MM.2018.112130359 }}</ref><ref name="Morris2017">{{cite web |last1=Morris |first1=John |title=Why Intel built a neuromorphic chip |url=https://www.zdnet.com/article/why-intel-built-a-neuromorphic-chip/ |website=ZDNet |access-date=17 August 2018 |language=en}}</ref><br />
<br />
<br />
比利时的微电子研究中心（IMEC）研发了世界上首个自学习神经形态芯片。这种基于丝状氧化物电阻存储技术 filamentary-oxide-based resistive memory technology（OxRAM）技术的大脑启发芯片具有自学习能力，并且已被证明具有创作音乐的能力。<ref name=":24">{{cite web |title=Imec demonstrates self-learning neuromorphic chip that composes music |url=https://www.imec-int.com/en/articles/imec-demonstrates-self-learning-neuromorphic-chip-that-composes-music |website=IMEC International |access-date=1 October 2019}}</ref>IMEC发布了由芯片原型机谱写的30秒旋律。向芯片加载一系列特征、风格统一的歌曲（古代比利时和法国长笛小步舞曲），芯片从中学习相关规则并将其应用于创作。<ref name=":25">{{cite web|last1=Bourzac|first1=Katherine|title=A Neuromorphic Chip That Makes Music|url=https://spectrum.ieee.org/a-neuromorphic-chip-that-makes-music|url-status=live|access-date=1 October 2019|website=IEEE Spectrum|date=May 23, 2017}}</ref><br />
<br />
<br />
由Henry Markram领导的蓝脑计划 The Blue Brain Project旨在建立小鼠大脑生理学细节的数字重建和模拟。蓝脑计划已经建立了啮齿动物大脑的电子模型，同时进行着尽可能多地复制其生理学细节的尝试。基于超级计算机的模拟为理解大脑的结构和功能提供了新的视角。<br />
<br />
<br />
欧盟资助了海德堡大学一系列促进BrainScaleS（神经形态混合系统中受大脑启发的多尺度计算）发展的项目，这是一台位于德国海德堡大学的混合模拟神经形态超级计算机。它是作为人类大脑计划中神经形态计算平台的一部分而开发的，是SpiNNaker超级计算机（基于数字技术）的补充。BrainScaleS中使用的体系架构模拟了生物神经元及其在物理层面上的连接；此外，由于这些组件是由硅制成的，这些模型神经元平均运行速度是生物神经元的864倍，这意味着在机器模拟中，24小时的实时时间仅为100秒。<ref name=":26">{{Cite web|date=2016-03-21|title=Beyond von Neumann, Neuromorphic Computing Steadily Advances|url=https://www.hpcwire.com/2016/03/21/lacking-breakthrough-neuromorphic-computing-steadily-advance/|access-date=2021-10-08|website=HPCwire|language=en-US}}</ref><br />
<br />
<br />
===神经形态传感器===<br />
神经形态系统的概念可以扩展到传感器（而不仅仅是计算单元）。一个用于检测光线的例子是类视网膜传感器，或者事件摄像机阵列。<br />
<br />
<br />
==伦理问题==<br />
虽然神经形态工程这个跨学科概念相对较新，但许多适用于类人机器 Human-like machines和人工智能的伦理讨论在神经形态系统领域也无法避免。另外，神经形态系统是为了模仿人类大脑而设计，这一底层逻辑也导致了一些新的伦理问题。<br />
<br />
<br />
然而，这种争论的实际情况是，神经形态硬件和人工“神经网络”是对大脑运作或处理信息过程的极其简化的模型，在大小和功能技术方面的复杂性相比而言比较低，在连接性方面也具有更加规则的结构。将神经形态芯片与大脑进行比较是一种非常粗糙的比较，类似于仅仅因为飞机有翅膀和尾巴就将它与鸟进行比较。事实上，当前最先进的人工智能在能耗效率和计算效率方面距离人脑神经认知系统仍有较大差距，而神经形态工程只是一种通过从大脑机制中获得灵感来缩小这种差距的尝试，就像许多工程设计中都具有生物启发的特征 Bio-inspired features一样。<br />
<br />
<br />
===公众担忧===<br />
由于公众认知的相关忧虑，神经形态工程学可能会受到严重的伦理限制。<ref name=":27">{{Cite journal|url=https://ai100.stanford.edu/sites/g/files/sbiybj9861/f/ai_100_report_0831fnl.pdf|title=Artificial Intelligence and Life in 2030|author=2015 Study Panel|date=September 2016|work=One Hundred Year Study on Artificial Intelligence (AI100)|publisher=Stanford University}}</ref>欧盟委员会进行的一项调查发现，60% 的欧盟公民希望禁止机器人参与照顾儿童、老人或残疾人的工作。此外，34% 的人支持禁止机器人用于教育，27% 的人支持禁止机器人用于医疗保健，20% 的人支持禁止机器人用于娱乐。欧盟委员会将以上领域划入“人类”范畴。报告指出，公众越来越关注能够模仿或复制人类行为的机器人。而神经形态工程，顾名思义，是为了模仿人脑的功能而设计的。<ref name=":1">{{Cite web|url=http://ec.europa.eu/commfrontoffice/publicopinion/archives/ebs/ebs_382_en.pdf|title=Special Eurobarometer 382: Public Attitudes Towards Robots|last=European Commission|date=September 2012|website=European Commission}}</ref><br />
<br />
<br />
围绕神经形态工程的公众担忧可能在未来变得更加严重。欧盟委员会发现，相比于55岁以上的欧盟公民，15至24岁的欧盟公民更有可能认为机器人像人（而不是像仪器）。当看到一张“类人”机器人的图片时，年龄在15岁至24岁之间的欧盟公民中有75% 的人表示这符合他们对机器人的想法，而55岁以上的欧盟公民中只有57% 的人有同样的反应。因此，神经形态系统可能因为其类似人类的特性而被归入许多欧盟公民希望在未来禁止使用的机器人类别。<ref name=":1" /><br />
<br />
<br />
===人格权问题===<br />
随着神经形态系统的日益发展，一些学者主张赋予这些系统人格权。如果是大脑赋予了人类人格，那么在多大程度上模仿人类大脑的神经形态系统才能被赋予人格权利？“人类大脑计划”旨在推进以大脑为灵感的计算机技术发展，该计划的批评者认为，神经形态计算机技术的进步可能导致机器意识或人格的形成。<ref name=":28">{{Cite journal|last=Aicardi|first=Christine|date=September 2018|title=Accompanying technology development in the Human Brain Project: From foresight to ethics management|journal=Futures|volume=102|pages=114–124|doi=10.1016/j.futures.2018.01.005|doi-access=free}}</ref>这些批评者认为，如果这些系统被当作人来对待，那么人类使用神经形态系统执行任务（包括终止神经形态系统）的行为，在道德上就可能是不被允许的，因为这些行为将违反神经形态系统的自主性。<ref name=":29">{{Cite journal|last=Lim|first=Daniel|date=2014-06-01|title=Brain simulation and personhood: a concern with the Human Brain Project|journal=Ethics and Information Technology|language=en|volume=16|issue=2|pages=77–89|doi=10.1007/s10676-013-9330-5|issn=1572-8439}}</ref><br />
<br />
<br><br />
<br />
==军民两用技术（军事应用）==<br />
'''联合人工智能中心 The Joint Artificial Intelligence Center'''（JAIC），是美国军队的一个分支，专门从事采购和实施用于战斗的人工智能软件和神经形态硬件。具体应用包括智能耳机、护目镜和机器人。JAIC打算高度依赖神经形态技术，使用神经形态技术来连接神经形态单位网络中的“每个战士、每个射手”。<br />
<br />
<br />
==法律问题==<br />
怀疑派认为，在法律上没有办法应用能够适用于神经形态技术的电子人格。在一封由285名法律、机器人技术、医学和伦理学专家的联名信中，作者们反对欧盟委员会提出的承认“智能机器人”为法人的提议。他们写道，“机器人的法律地位无法从自然人模型中推导出来，因为机器人将被赋予人权，如尊严权、完整权、报酬权或公民权，从而直接面临人权问题。这将有悖于《欧联基本权利宪章》和《欧洲保障人权和根本自由公约》”。<ref name=":30">{{Cite web|url=http://www.robotics-openletter.eu/|title=Robotics Openletter {{!}} Open letter to the European Commission|language=fr-FR|access-date=2019-05-10}}</ref><br />
<br />
<br />
===所有权及财产权问题===<br />
法律界围绕财产权和人工智能有着重大争论。在Acohs Pty有限公司诉Ucorp Pty有限公司一案中，澳大利亚联邦法院的Christopher Jessup法官发现，版权保护不适用于材料安全数据表的源代码，因为它是由软件界面生成而非人类工作者生成的。<ref name=":31">{{Cite web|url=http://www.lavan.com.au/advice/intellectual_property_technology/copyright_in_source_code_and_digital_products|title=Copyright in source code and digital products|last=Lavan|website=Lavan|language=en|access-date=2019-05-10}}</ref>同样的问题可能也适用于神经形态系统：如果一个神经形态系统成功地模仿了人类的大脑，并产生了一部原创作品，那么该如何确认这部作品的所有权归属？<ref name=":32">{{cite journal |last1=Eshraghian|first1=Jason K. |title=Human Ownership of Artificial Creativity |journal=Nature Machine Intelligence |date=9 March 2020 |volume=2 |pages=157–160 |doi=10.1038/s42256-020-0161-x}}</ref><br />
<br />
<br />
==神经忆阻系统==<br />
神经忆阻系统是神经形态计算系统的一个亚类，主要研究利用忆阻器 Memristors实现神经可塑性 Neuroplasticity。神经形态工程的重点是模拟生物行为，而神经忆阻系统的重点是提取。<ref name=":33">{{Cite web|url=https://digitalops.sandia.gov/Mediasite/Play/a10cf6ceb55d47608bb8326dd00e46611d|title=002.08 N.I.C.E. Workshop 2014: Towards Intelligent Computing with Neuromemristive Circuits and Systems - Feb. 2014|website=digitalops.sandia.gov|access-date=2019-08-26}}</ref>举个例子，一个神经忆阻系统可能用抽象的神经网络模型替代皮层微电路的行为细节。<ref name=":34">C. Merkel and D. Kudithipudi, "Neuromemristive extreme learning machines for pattern classification," ISVLSI, 2014.</ref><br />
<br />
<br />
受神经元启发、使用忆阻器实现的阈值逻辑函数<ref name="Maan 1–13" />在高级模式识别中有着广泛的应用，最近报道中其应用包括语音识别<ref name=":35">{{Cite journal|title = Memristor pattern recogniser: isolated speech word recognition|journal = Electronics Letters|pages = 1370–1372|volume = 51|issue = 17|doi = 10.1049/el.2015.1428|first1 = A.K.|last1 = Maan|first2 = A.P.|last2 = James|first3 = S.|last3 = Dimitrijev|year = 2015|bibcode = 2015ElL....51.1370M|hdl = 10072/140989|url = https://semanticscholar.org/paper/48d3ab11ec6e213b62f11eedcfb7b7febb058674|hdl-access = free}}</ref>、人脸识别<ref name=":36">{{Cite journal|title = Memristive Threshold Logic Face Recognition|journal = Procedia Computer Science|date = 2014-01-01|pages = 98–103|volume = 41|series = 5th Annual International Conference on Biologically Inspired Cognitive Architectures, 2014 BICA|doi = 10.1016/j.procs.2014.11.090|first1 = Akshay Kumar|last1 = Maan|first2 = Dinesh S.|last2 = Kumar|first3 = Alex Pappachen|last3 = James|doi-access = free}}</ref>和物体识别<ref name=":37">{{Cite journal|title = Memristive Threshold Logic Circuit Design of Fast Moving Object Detection|journal = IEEE Transactions on Very Large Scale Integration (VLSI) Systems|date = 2015-10-01|issn = 1063-8210|pages = 2337–2341|volume = 23|issue = 10|doi = 10.1109/TVLSI.2014.2359801|first1 = A.K.|last1 = Maan|first2 = D.S.|last2 = Kumar|first3 = S.|last3 = Sugathan|first4 = A.P.|last4 = James|arxiv = 1410.1267}}</ref>。阈值逻辑函数还可以用来取代传统的数字逻辑门。<ref name=":38">{{Cite journal|title = Resistive Threshold Logic|journal = IEEE Transactions on Very Large Scale Integration (VLSI) Systems|date = 2014-01-01|issn = 1063-8210|pages = 190–195|volume = 22|issue = 1|doi = 10.1109/TVLSI.2012.2232946|first1 = A.P.|last1 = James|first2 = L.R.V.J.|last2 = Francis|first3 = D.S.|last3 = Kumar|arxiv = 1308.0090}}</ref><ref name=":39">{{Cite journal|title = Threshold Logic Computing: Memristive-CMOS Circuits for Fast Fourier Transform and Vedic Multiplication|journal = IEEE Transactions on Very Large Scale Integration (VLSI) Systems|date = 2015-11-01|issn = 1063-8210|pages = 2690–2694|volume = 23|issue = 11|doi = 10.1109/TVLSI.2014.2371857|first1 = A.P.|last1 = James|first2 = D.S.|last2 = Kumar|first3 = A.|last3 = Ajayan|arxiv = 1411.5255}}</ref><br />
<br />
<br />
对于理想的无源记忆电路，电路的内部记忆可以用精确的方程（Caravelli-Traversa-Di Ventra方程) 来描述:<ref name=":40">{{cite journal |last=Caravelli |display-authors=etal|arxiv=1608.08651 |title=The complex dynamics of memristive circuits: analytical results and universal slow relaxation |year=2017 |doi=10.1103/PhysRevE.95.022140 |pmid= 28297937 |volume=95 |issue= 2 |pages= 022140 |journal=Physical Review E|bibcode=2017PhRvE..95b2140C}}</ref><br />
<br />
<br />
:<math> \frac{d}{dt} \vec{W} = \alpha \vec{W}-\frac{1}{\beta} (I+\xi \Omega W)^{-1} \Omega \vec S </math><br />
<br />
<br />
Caravelli-Traversa-Di Ventra方程是描述物理记忆网络和外部源性质的函数。在上述方程中，<math>\alpha</math>是“遗忘”时间尺度常数，<math>\xi=r-1</math>，<math>r =\frac{R\text_{off}}{R_\text{on}}</math>是记忆电阻器off状态和on状态极限电阻值之比，<math>\vec S</math>是电路源的矢量，<math>\Omega</math>是电路基本环路的投影。常数<math>\beta</math>具有电压的量纲，与记忆电阻器的特性有关；其物理原型是导体中的电荷迁移率。对角矩阵和向量 <math>W=\operatorname{diag}(\vec W)</math>和<math>\vec W</math> 分别是忆阻器的内阻，值在0到1之间。因此，这个等式需要在内存值=上添加额外约束以保证可靠性。<br />
<br />
<br><br />
<br />
==另见==<br />
{{Columns-list|colwidth=18em|<br />
*[[AI加速器(计算机硬件)|AI加速器]] <br />
*[[人工大脑]] <br />
*[[生物形态的]] <br />
*[[认知计算机]] <br />
*[[计算与神经系统]] <br />
*[[事件相机]] <br />
* [[神经机器人学]] <br />
*[[光流传感器]] <br />
*[[物理神经网络]] <br />
*[[突触]] <br />
*[[Retinomorphic传感器]] <br />
*[[视觉芯片]] <br />
*[[视觉处理单元]] <br />
*[[人工智能硬件]] <br />
}}<br />
<br />
<br />
==参考文献==<br />
<references /><br />
<br />
<br><br />
<br />
==外部链接==<br />
<br />
*[https://web.archive.org/web/20150727034331/http://ine-web.org/workshops/workshops-overview 碲化物神经形态工程工作室]<br />
*[https://archive.today/20130115190057/http://capocaccia.ethz.ch/ CapoCaccia认知神经形态工程工作室]<br />
*[https://web.archive.org/web/20190716132350/http://www.ine-web.org/ 神经形态工程研究所]<br />
*[http://www.ine-news.org/ INE新闻站点]<br />
*[http://www.frontiersin.org/neuromorphic_engineering 神经形态工程学前沿]<br />
*[http://www.cns.caltech.edu/ 加州理工学院计算与神经系统系]].<br />
*[http://www.humanbrainproject.eu/ 人脑项目官方网站]<br />
*[https://www.the-scientist.com/features/building-a-silicon-brain-65738 打造硅脑: 基于生物神经元的计算机芯片可能有助于模拟更大、更复杂的大脑模型]]<br />
<br />
<br />
==编者推荐==<br />
===集智课程===<br />
====[https://campus.swarma.org/course/4370 神经动力学模型读书会]====<br />
人类大脑是一个由数以百亿计的神经元相互连接所构成的复杂系统。自神经科学这一学科建立以来，从解析神经元之间的相互作用机理、到刻画皮层柱之间的连接形式、再到探究脑区间不同认知功能的分离与整合模式，无数科学家试图从不同尺度研究大脑，以期揭示人脑这一最为复杂的神经系统的工作模式，进而理解语言、情绪、记忆和社会交往等高级认知活动的底层神经机制，并一定程度上启发通用人工智能机器人的设计。<br />
<br />
近年来，脱胎于系统科学的动力学建模方法，逐渐被广泛地应用于神经科学研究中，其作为一种绝佳的数理工具，愈发地受到研究人员的重视，在类脑计算、脑认知原理解析和脑重大疾病致病机理探索等具体方面，发挥着不可替代的作用。<br />
<br />
本着促进神经科学、系统科学以及计算机科学等不同领域的学术工作者的交流与合作，激发有志于加入脑科学与类脑研究这一领域的同学们的研究兴趣，来自国内外多所知名高校的专家学者（详见发起人介绍）共同发起了「神经动力学模型」读书会，聚焦于神经科学中的动力学建模这一前沿课题，讨论他们在研究一线中遇到的实际困惑，对相关文献进行深入梳理、激发跨学科的学术火花。<br />
<br />
<br />
====[https://campus.swarma.org/course/1925 廖旭红：从脑网络到认知]====<br />
脑是自然界最为复杂的系统之一。近年来，研究者逐渐意识到脑网络是人脑进行高效信息传输、实现各种认知功能的基础，开始全面而细致刻画人脑结构和功能连接模式。该报告将从基于磁共振成像技术和复杂网络理论的脑网络研究框架出发，探讨人脑结构和功能的组织原则以及认知行为的脑机制。<br />
<br />
<br />
====[https://campus.swarma.org/course/951 李印赟：神经突触可塑性微观机制及神经环路研究]====<br />
该讲座分为两部分：第一部分为神经元的轴突输运特征及其与神经形态之间的关系，第二部分为神经突触可塑性中涉及的囊泡输运和释放，短时突触可塑性的神经细胞差异性及其在神经元微环路中细胞发放特征的作用。<br />
<br />
<br />
----<br />
本中文词条由Shenky20翻译，[[用户:薄荷|薄荷]]编辑，如有问题，欢迎在讨论页面留言。<br />
<br />
<br />
<br />
'''本词条内容源自wikipedia及公开资料，遵守 CC3.0协议。'''<br />
<br />
[[Category:电气工程]]<br />
[[Category:神经科学]]<br />
[[Category:人工智能]]<br />
[[Category:机器人]]</div>Shenky20https://wiki.swarma.org/index.php?title=%E7%A5%9E%E7%BB%8F%E5%BD%A2%E6%80%81%E8%AE%A1%E7%AE%97&diff=30765神经形态计算2022-04-29T03:23:13Z<p>Shenky20：</p>
<hr />
<div><br />
此词条由神经动力学模型读书会词条梳理志愿者Shenky20翻译审校，未经专家审校，带来阅读不便，请见谅。<br />
<br />
{{Use American English|date = January 2019}}<br />
{{Use mdy dates|date = January 2019}}<br />
<br />
'''Neuromorphic engineering''', also known as '''neuromorphic computing''',<ref name=":3">{{Cite journal | doi = 10.1145/2601069| title = Neuromorphic computing gets ready for the (really) big time| journal = [[Communications of the ACM]]| volume = 57| issue = 6| year = 2014| last1 = Monroe | first1 = D. | pages = 13–15| s2cid = 20051102}}</ref><ref name=":4">{{Cite journal | doi = 10.1088/0957-4484/21/17/175202| pmid = 20368686| title = Nanotube devices based crossbar architecture: Toward neuromorphic computing| journal = Nanotechnology| volume = 21| issue = 17| pages = 175202| year = 2010| last1 = Zhao | first1 = W. S. | last2 = Agnus | first2 = G. | last3 = Derycke | first3 = V. | last4 = Filoramo | first4 = A. | last5 = Bourgoin | first5 = J. -P. | last6 = Gamrat | first6 = C. | bibcode = 2010Nanot..21q5202Z| url = https://zenodo.org/record/3428659}}</ref><ref name="humanbrainproject">{{YouTube|id=6RoiZ90mGfw|title=The Human Brain Project SP 9: Neuromorphic Computing Platform}}</ref> is the use of [[very-large-scale integration]] (VLSI) systems containing electronic [[analog circuit]]s to mimic neuro-biological architectures present in the nervous system. A neuromorphic computer/chip is any device that uses physical artificial neurons (made from silicon) to do computations.<ref name=":5">{{cite journal|last1=Mead|first1=Carver|title=Neuromorphic electronic systems|journal=Proceedings of the IEEE|date=1990|volume=78|issue=10|pages=1629–1636|doi=10.1109/5.58356|url=https://authors.library.caltech.edu/53090/1/00058356.pdf}}</ref><ref name=":2" /> In recent times, the term ''neuromorphic'' has been used to describe analog, digital, mixed-mode analog/digital VLSI, and software systems that implement models of [[neural system]]s (for [[perception]], [[motor control]], or [[multisensory integration]]). The implementation of neuromorphic computing on the hardware level can be realized by oxide-based [[memristor]]s,<ref name="Maan 1–13">{{Cite journal|last1=Maan|first1=A. K.|last2=Jayadevi|first2=D. A.|last3=James|first3=A. P.|date=2016-01-01|title=A Survey of Memristive Threshold Logic Circuits|journal=IEEE Transactions on Neural Networks and Learning Systems|volume=PP|issue=99|pages=1734–1746|doi=10.1109/TNNLS.2016.2547842|pmid=27164608|issn=2162-237X|arxiv=1604.07121|bibcode=2016arXiv160407121M|s2cid=1798273}}</ref> spintronic memories, threshold switches, and [[transistor]]s.<ref name=":6">{{Cite journal|title = Mott Memory and Neuromorphic Devices|journal = Proceedings of the IEEE|date = 2015-08-01|issn = 0018-9219|pages = 1289–1310|volume = 103|issue = 8|doi = 10.1109/JPROC.2015.2431914|first1 = You|last1 = Zhou|first2 = S.|last2 = Ramanathan|s2cid = 11347598|url = https://zenodo.org/record/895565}}</ref><ref name=":2">{{Cite document|title=Neuromorphic Circuits With Neural Modulation Enhancing the Information Content of Neural Signaling {{!}} International Conference on Neuromorphic Systems 2020|language=EN|doi=10.1145/3407197.3407204|s2cid=220794387}}</ref> Training software-based neuromorphic systems of spiking neural networks can be achieved using error backpropagation, e.g., using [[Python (programming language)|Python]] based frameworks such as snnTorch,<ref name=":7">{{cite journal |last1=Eshraghian|first1=Jason K.|last2=Ward|first2=Max|last3=Neftci |first3=Emre|last4=Wang|first4=Xinxin|last5=Lenz|first5=Gregor|last6=Dwivedi|first6=Girish|last7=Bennamoun|first7=Mohammed|last8=Jeong|first8=Doo Seok|last9=Lu|first9=Wei D.|title=Training Spiking Neural Networks Using Lessons from Deep Learning |date=1 October 2021 |arxiv=2109.12894 }}</ref> or using canonical learning rules from the biological learning literature, e.g., using BindsNet.<ref name=":8">{{Cite web | url=https://github.com/Hananel-Hazan/bindsnet | title=Hananel-Hazan/bindsnet: Simulation of spiking neural networks (SNNs) using PyTorch.| date=31 March 2020}}</ref><br />
<br />
<br />
'''<font color="#ff8000">神经形态工程Neuromorphic engineering</font>'''（也称为'''<font color="#ff8000">神经形态计算Neuromorphic computing</font>'''或'''<font color="#ff8000">类脑计算Brain-inspired computing</font>'''）<ref name=":3" /><ref name=":4" /><ref name="humanbrainproject" />是指使用包含电子'''<font color="#ff8000">模拟电路Analog circuit</font>'''的'''<font color="#ff8000">超大规模集成电路Very-large-scale integration</font>'''系统来模拟神经系统中生理结构的研究方法。神经形态计算机或神经形态芯片包括任何使用由硅制成的人造神经元进行计算的设备。<ref name=":5" /><ref name=":2" />近年来，神经形态学（neuromorphic）这个术语被用来描述能够实现'''<font color="#ff8000">神经系统Neural system</font>'''模型功能(如'''<font color="#ff8000">感知Perception</font>'''、'''<font color="#ff8000">运动控制Motor control</font>'''，'''<font color="#ff8000">多感官整合Multisensory integration</font>'''等)的模拟、数字、模拟/数字混合模式超大规模集成电路和软件系统。神经形态计算的硬件实现可以通过基于氧化物的'''<font color="#ff8000">记忆电阻器Memristor</font>'''（简称忆阻器）、<ref name="Maan 1–13" />自旋电子存储器、阈值开关和'''<font color="#ff8000">晶体管Transistor</font>'''来实现。<ref name=":6" /><ref name=":2" />对基于软件的脉冲神经网络系统的训练可以通过误差反向传播机制来实现，例如，使用snnTorch等基于Python的框架，<ref name=":7" />或使用BindsNet等典型的受生物启发的学习模式。<ref name=":8" /><br />
<br />
A key aspect of neuromorphic engineering is understanding how the morphology of individual neurons, circuits, applications, and overall architectures creates desirable computations, affects how information is represented, influences robustness to damage, incorporates learning and development, adapts to local change (plasticity), and facilitates evolutionary change.<br />
<br />
<br />
神经形态工程领域的一个关键问题，就是理解'''<font color="#32CD32">单个神经元形态、神经回路、应用和整体结构</font>'''如何产生理想的计算，如何影响信息的表示和对破坏的鲁棒性，如何整合学习和发展，如何适应局部变化(可塑性) 并促进逐渐发展的变化。<br />
<br />
Neuromorphic engineering is an interdisciplinary subject that takes inspiration from [[biology]], [[physics]], [[mathematics]], [[computer science]], and [[electronic engineering]]<ref name=":2" /> to design artificial neural systems, such as [[Machine vision|vision systems]], head-eye systems, auditory processors, and autonomous robots, whose physical architecture and design principles are based on those of biological nervous systems.<ref name=":9">{{Cite journal | doi = 10.1155/2012/705483| title = Qualitative Functional Decomposition Analysis of Evolved Neuromorphic Flight Controllers| journal = Applied Computational Intelligence and Soft Computing| volume = 2012| pages = 1–21| year = 2012| last1 = Boddhu | first1 = S. K. | last2 = Gallagher | first2 = J. C. | doi-access = free}}</ref> It was developed by [[Carver Mead]]<ref name=":10">{{cite web|title=carver mead website|last1=Mead|first1=Carver|url=http://www.carvermead.caltech.edu/index.html|website=carvermead}}</ref> in the late 1980s.<br />
<br />
<br />
神经形态工程是以'''<font color="#ff8000">生物学Biology</font>'''、'''<font color="#ff8000">物理学Physics</font>'''、'''<font color="#ff8000">数学Mathematics</font>'''、'''<font color="#ff8000">计算机科学Computer science</font>'''和'''<font color="#ff8000">电子工程Electronic engineering<ref name=":2" /></font>'''等学科为基础，设计人工神经系统（如'''<font color="#ff8000">视觉系统Vision systems</font>'''、头眼系统、听觉处理器以及物理结构和设计原则都受启发于生物神经系统的自主机器人）的一门交叉学科。<ref name=":9" />20世纪80年代后期，卡弗·米德极大地推动了神经形态工程领域的发展。<ref name=":10" /><br />
<br />
==Neurological inspiration ==<br />
==神经生物学启发 ==<br />
Neuromorphic engineering is set apart by the inspiration it takes from what we know about the structure and operations of the [[brain]]. Neuromorphic engineering translates what we know about the brain's function into computer systems. Work has mostly focused on replicating the analog nature of [[biological computation]] and the role of [[neuron]]s in [[cognition]].<br />
<br />
<br />
神经形态工程形成的灵感来源于现已知的'''<font color="#ff8000">大脑Brain</font>'''结构及运作机制知识，它将我们对大脑功能的了解用于研究和优化计算机系统。该领域工作主要集中于对'''<font color="#ff8000">生物计算Biological computation</font>'''模拟特性和'''<font color="#ff8000">神经元Neuron</font>'''在'''<font color="#ff8000">认知Cognition</font>'''中所发挥作用的复刻。<br />
<br />
The biological processes of neurons and their [[synapse]]s are dauntingly complex, and thus very difficult to artificially simulate. A key feature of biological brains is that all of the processing in neurons use analog chemical signals. This makes it hard to replicate brains in computers because the current generation of computers is completely digital. However, the characteristics of these parts can be abstracted into mathematical functions that closely capture the essence of the neuron's operations.<br />
<br />
神经元及其'''<font color="#ff8000">突触Synapse</font>'''的生理学过程极其复杂，难以进行人工模拟。神经元在所有处理过程中都使用化学模拟信号，这是大脑的一个关键生理学特征。这个特征大大增加了在计算机中复制大脑的难度，因为目前的计算机是完全数字化的。然而，部分特征可以抽象为数学函数，这些函数能够紧密捕捉神经元运行的本质。<br />
<br />
The goal of neuromorphic computing is not to perfectly mimic the brain and all of its functions, but instead to extract what is known of its structure and operations to be used in a practical computing system. No neuromorphic system will claim nor attempt to reproduce every element of neurons and synapses, but all adhere to the idea that computation is highly [[distributed processing|distributed]] throughout a series of small computing elements analogous to a neuron. While this sentiment is standard, researchers chase this goal with different methods.<ref name=":11">{{Cite journal | doi = 10.1088/1741-2560/13/5/051001| title = Large-scale neuromorphic computing systems| journal = Journal of Neural Engineering| volume = 13| pages = 1–15| year = 2016| last1 = Furber | first1 = Steve| issue = 5| pmid = 27529195| bibcode = 2016JNEng..13e1001F| doi-access = free}}</ref><br />
<br />
神经形态计算的目标不是完美地模拟大脑及其所有功能，而是利用已知的大脑结构和运转机制来研发或优化实际的计算系统。任何神经形态学系统都不会声称或试图复制神经元和突触中的每一个元素，但所有人都一致认可将计算高度'''<font color="#ff8000">分散Distribute</font>'''于一系列类似于神经元的小型计算元素的理念。研究人员用不同的方法来追求这一普遍目标。<ref name=":11" /><br />
<br />
<br />
==Examples==<br />
==代表性成果==<br />
As early as 2006, researchers at [[Georgia Tech]] published a field programmable neural array.<ref name=":12">{{Cite book|title = A field programmable neural array|last1 = Farquhar|first1 = Ethan|date = May 2006|journal = IEEE International Symposium on Circuits and Systems|pages = 4114–4117|last2 = Hasler|first2 = Paul.|doi = 10.1109/ISCAS.2006.1693534|isbn = 978-0-7803-9389-9|s2cid = 206966013}}</ref> This chip was the first in a line of increasingly complex arrays of floating gate transistors that allowed programmability of charge on the gates of [[MOSFET]]s to model the channel-ion characteristics of neurons in the brain and was one of the first cases of a silicon programmable array of neurons.<br />
<br />
早在2006年，佐治亚理工学院的研究人员就研发出了一种现场可编程神经阵列。<ref name=":12" />在此之后，一系列越来越复杂的浮栅晶体管阵列被成功研发出来，这些晶体管阵列可以通过在'''<font color="#ff8000">金属-氧化物半导体效应晶体管MOSFET</font>'''的栅极上编程来模拟大脑中神经元的离子通道特性，这也是以硅为主要材料的可编程神经元阵列的最早成功案例之一。<br />
<br />
In November 2011, a group of [[MIT]] researchers created a computer chip that mimics the analog, ion-based communication in a synapse between two neurons using 400 transistors and standard [[CMOS]] manufacturing techniques.<ref name=":13">{{cite web|title=MIT creates "brain chip"|url=http://www.extremetech.com/extreme/105067-mit-creates-brain-chip|access-date=4 December 2012}}</ref><ref name="Neuromorphic silicon paper">{{cite journal|title=Neuromorphic silicon neurons and large-scale neural networks: challenges and opportunities|doi=10.3389/fnins.2011.00108|pmid=21991244|pmc=3181466|volume=5|pages=108|journal=Frontiers in Neuroscience|year=2011|last1=Poon|first1=Chi-Sang|last2=Zhou|first2=Kuan|doi-access=free}}</ref><br />
<br />
2011年11月，麻省理工学院的一组研究人员研发出一种计算机芯片，该芯片上使用标准的'''<font color="#ff8000">互补金属氧化物半导体CMOS</font>'''制造技术集成了400个晶体管，能够模拟神经元间突触中基于离子的通讯。<ref name=":13" /><ref name="Neuromorphic silicon paper" /><br />
<br />
In June 2012, [[spintronic]] researchers at [[Purdue University]] presented a paper on the design of a neuromorphic chip using [[Spin valve|lateral spin valve]]s and [[memristor]]s. They argue that the architecture works similarly to neurons and can therefore be used to test methods of reproducing the brain's processing. In addition, these chips are significantly more energy-efficient than conventional ones.<ref name="Spin Devices Prop">{{Cite arXiv|title=Proposal For Neuromorphic Hardware Using Spin Devices|eprint=1206.3227|last1=Sharad|first1=Mrigank|last2=Augustine|first2=Charles|last3=Panagopoulos|first3=Georgios|last4=Roy|first4=Kaushik|class=cond-mat.dis-nn|year=2012}}</ref><br />
<br />
2012年6月，普渡大学的'''<font color="#ff8000">自旋电子学Spintronic</font>'''研究人员发表了一篇关于利用'''<font color="#ff8000">侧向自旋阀Lateral spin valves</font>'''和'''<font color="#ff8000">忆阻器Memristors</font>'''设计神经形态芯片的论文。他们认为，这种芯片结构的工作原理与神经元相似，因此可以用于大脑运行机制的复刻方法的测试。此外，这些芯片在能耗方面明显优于传统芯片。<ref name="Spin Devices Prop" /><br />
<br />
Research at [[HP Labs]] on Mott memristors has shown that while they can be non-[[Volatile memory|volatile]], the volatile behavior exhibited at temperatures significantly below the [[phase transition]] temperature can be exploited to fabricate a [[neuristor]],<ref name=":0" /> a biologically-inspired device that mimics behavior found in neurons.<ref name=":0">{{Cite journal | doi = 10.1038/nmat3510| pmid = 23241533| title = A scalable neuristor built with Mott memristors| journal = Nature Materials| volume = 12| issue = 2| pages = 114–7| year = 2012| last1 = Pickett | first1 = M. D. | last2 = Medeiros-Ribeiro | first2 = G. | last3 = Williams | first3 = R. S. | bibcode = 2013NatMa..12..114P| s2cid = 16271627| url = https://semanticscholar.org/paper/b6ba6f496ace2b947f111059663e76bb60e9efeb}}</ref> In September 2013, they presented models and simulations that show how the spiking behavior of these neuristors can be used to form the components required for a [[Turing machine]].<ref name=":14">{{cite journal|doi=10.1088/0957-4484/24/38/384002|title=Phase transitions enable computational universality in neuristor-based cellular automata|author1=Matthew D Pickett|author2=R Stanley Williams|name-list-style=amp|date=September 2013|publisher=IOP Publishing Ltd|journal=Nanotechnology|volume=24|issue=38|pmid=23999059|bibcode=2013Nanot..24L4002P|at=384002}}</ref><br />
<br />
'''<font color="#ff8000">惠普实验室HP labs</font>'''在莫特忆阻器上的研究表明，尽管它们可以是非'''<font color="#ff8000">易失性Volatile</font>'''的，但是在'''<font color="#ff8000">相变Phase transition</font>'''温度以下时表现出的易失性行为可以被用来制造'''<font color="#ff8000">类神经元电阻器Neuristor</font>'''（一种生物学启发的模仿神经元行为的硬件）<ref name=":0" />。2013年9月，他们通过模型和仿真展示了这些类神经元电阻器的脉冲行为如何产生'''<font color="#ff8000">图灵机Turing machine</font>'''的所需元素。<ref name=":14" /><br />
<br />
[[Neurogrid]], built by ''Brains in Silicon'' at [[Stanford University]],<ref name=":15">{{cite journal|last1=Boahen|first1=Kwabena|title=Neurogrid: A Mixed-Analog-Digital Multichip System for Large-Scale Neural Simulations|journal=Proceedings of the IEEE|date=24 April 2014|volume=102|issue=5|pages=699–716|doi=10.1109/JPROC.2014.2313565|s2cid=17176371}}</ref> is an example of hardware designed using neuromorphic engineering principles. The circuit board is composed of 16 custom-designed chips, referred to as NeuroCores. Each NeuroCore's analog circuitry is designed to emulate neural elements for 65536 neurons, maximizing energy efficiency. The emulated neurons are connected using digital circuitry designed to maximize spiking throughput.<ref name=":16">{{cite journal|doi=10.1038/503022a|pmid = 24201264|title = Neuroelectronics: Smart connections|journal = Nature|volume = 503|issue = 7474|pages = 22–4|year = 2013|last1 = Waldrop|first1 = M. Mitchell|bibcode = 2013Natur.503...22W|doi-access = free}}</ref><ref name=":17">{{cite journal|doi=10.1109/JPROC.2014.2313565|title = Neurogrid: A Mixed-Analog-Digital Multichip System for Large-Scale Neural Simulations|journal = Proceedings of the IEEE|volume = 102|issue = 5|pages = 699–716|year = 2014|last1 = Benjamin|first1 = Ben Varkey|last2 = Peiran Gao|last3 = McQuinn|first3 = Emmett|last4 = Choudhary|first4 = Swadesh|last5 = Chandrasekaran|first5 = Anand R.|last6 = Bussat|first6 = Jean-Marie|last7 = Alvarez-Icaza|first7 = Rodrigo|last8 = Arthur|first8 = John V.|last9 = Merolla|first9 = Paul A.|last10 = Boahen|first10 = Kwabena|s2cid = 17176371}}</ref><br />
<br />
'''<font color="#ff8000">神经栅格Neurogrid<ref name=":15" /></font>'''是由斯坦福大学Brains in Silicon公司研发的、使用神经形态工程原理设计的硬件。该电路板由16个定制设计的芯片组成（NeuroCores）。在设计中，每个NeuroCore芯片的模拟电路对65536个神经元的神经元素进行模拟，以最大限度地提高能量效率。模拟出的神经元通过设计的数字电路连接，以最大化脉冲吞吐量。<ref name=":16" /><ref name=":17" /><br />
<br />
A research project with implications for neuromorphic engineering is the [[Human Brain Project]] that is attempting to simulate a complete human brain in a supercomputer using biological data. It is made up of a group of researchers in neuroscience, medicine, and computing.<ref name=":18">{{cite web|title=Involved Organizations|url=http://www.humanbrainproject.eu/partners.html|access-date=22 February 2013|url-status=dead|archive-url=https://web.archive.org/web/20130302142627/http://www.humanbrainproject.eu/partners.html|archive-date=2 March 2013}}</ref> [[Henry Markram]], the project's co-director, has stated that the project proposes to establish a foundation to explore and understand the brain and its diseases, and to use that knowledge to build new computing technologies. The three primary goals of the project are to better understand how the pieces of the brain fit and work together, to understand how to objectively diagnose and treat brain diseases, and to use the understanding of the human brain to develop neuromorphic computers. That the simulation of a complete human brain will require a supercomputer a thousand times more powerful than today's encourages the current focus on neuromorphic computers.<ref name=":19">{{cite web|title=Human Brain Project|url=http://www.humanbrainproject.eu|access-date=22 February 2013}}</ref> $1.3 billion has been allocated to the project by The [[European Commission]].<ref name=":20">{{cite web|title=The Human Brain Project and Recruiting More Cyberwarriors|url=http://www.marketplace.org/topics/tech/human-brain-project-and-recruiting-more-cyberwarriors|access-date=22 February 2013|date=January 29, 2013}}</ref><br />
<br />
'''<font color="#ff8000">人类大脑计划Human Brain Project</font>'''对神经形态工程具有较大影响，其主要任务是尝试用生物数据在超级计算机中模拟完整的人脑。人类大脑计划由神经科学、医学和计算机科学背景的研究人员组成。<ref name=":18" />该项目的联合主管亨利•马克拉姆（Henry Markram）表示，人类大脑计划的目的是建立一个探索和了解脑科学和脑疾病知识的基础，并利用这些知识来构建更先进的计算机技术。这个项目的三个主要目标分别是: 更好地理解大脑的各个部分是如何相互配合协同工作的; 理解如何客观地诊断和治疗脑部疾病; 以及利用对人类大脑的理解来开发神经形态计算机。模拟一个完整的人类大脑需要一台比现在强大一千倍的超级计算机，这不断激发着对神经形态计算机领域的研究兴趣。<ref name=":19" />欧盟委员会已经向人类大脑计划拨款13亿美元。<ref name=":20" /><br />
<br />
Other research with implications for neuromorphic engineering involves the [[BRAIN Initiative]]<ref name="economist">[https://www.economist.com/news/science-and-technology/21582495-computers-will-help-people-understand-brains-better-and-understanding-brains Neuromorphic computing: The machine of a new soul], The Economist, 2013-08-03</ref> and the [[TrueNorth]] chip from [[IBM]].<ref name=":21">{{cite journal|last1=Modha|first1=Dharmendra|title=A million spiking-neuron integrated circuit with a scalable communication network and interface|journal=Science|date=Aug 2014|volume=345|issue=6197|pages=668–673|doi=10.1126/science.1254642|pmid=25104385|bibcode=2014Sci...345..668M|s2cid=12706847}}</ref> Neuromorphic devices have also been demonstrated using nanocrystals, nanowires, and conducting polymers.<ref name=":22">{{Cite web|url=http://jessamynfairfield.com/wp-content/uploads/2017/03/PWMar17Fairfield.pdf|title=Smarter Machines|last=Fairfield|first=Jessamyn|date=March 1, 2017}}</ref><br />
<br />
其他与神经形态工程有关的研究还包括'''<font color="#ff8000">脑计划BRAIN initiative</font>'''，<ref name="economist" />和IBM研发的'''<font color="#ff8000">TrueNorth</font>'''芯片。<ref name=":21" />使用纳米晶体、纳米线和导电聚合物也能够用于制造神经形态学硬件。<ref name=":22" /><br />
<br />
[[Intel]] unveiled its neuromorphic research chip, called “[[Intel Loihi|Loihi]]”, in October 2017. The chip uses an asynchronous [[spiking neural network]] (SNN) to implement adaptive self-modifying event-driven fine-grained parallel computations used to implement learning and inference with high efficiency.<ref name=":23">{{cite journal |last1=Davies |first1=Mike |title=Loihi: A Neuromorphic Manycore Processor with On-Chip Learning |journal=IEEE Micro |date=January 16, 2018 |volume=38 |issue=1 |pages=82–99 |display-authors=etal|doi=10.1109/MM.2018.112130359 |s2cid=3608458 }}</ref><ref name="Morris2017">{{cite web |last1=Morris |first1=John |title=Why Intel built a neuromorphic chip |url=https://www.zdnet.com/article/why-intel-built-a-neuromorphic-chip/ |website=ZDNet |access-date=17 August 2018 |language=en}}</ref><br />
<br />
2017年10月，英特尔发布了神经形态芯片'''<font color="#ff8000">Loihi</font>'''。该芯片采用异步'''<font color="#ff8000">脉冲神经网络Spiking neural network</font>'''实现了自适应、自修改、事件驱动的细粒度并行计算，实现了高效的学习和推理。<ref name=":23" /><ref name="Morris2017" /><br />
<br />
[[IMEC]], a Belgium-based nanoelectronics research center, demonstrated the world's first self-learning neuromorphic chip. The brain-inspired chip, based on OxRAM technology, has the capability of self-learning and has been demonstrated to have the ability to compose music.<ref name=":24">{{cite web |title=Imec demonstrates self-learning neuromorphic chip that composes music |url=https://www.imec-int.com/en/articles/imec-demonstrates-self-learning-neuromorphic-chip-that-composes-music |website=IMEC International |access-date=1 October 2019}}</ref> IMEC released the 30-second tune composed by the prototype. The chip was sequentially loaded with songs in the same time signature and style. The songs were old Belgian and French flute minuets, from which the chip learned the rules at play and then applied them.<ref name=":25">{{cite web|last1=Bourzac|first1=Katherine|title=A Neuromorphic Chip That Makes Music|url=https://spectrum.ieee.org/a-neuromorphic-chip-that-makes-music|url-status=live|access-date=1 October 2019|website=IEEE Spectrum|date=May 23, 2017}}</ref><br />
<br />
比利时的'''<font color="#ff8000">微电子研究中心IMEC</font>'''研发了世界上首个自学习神经形态芯片。这种基于'''<font color="#32CD32"> OxRAM(filamentary-oxide-based resistive memory technology)</font>''' 技术的大脑启发芯片具有自学习能力，并且已被证明具有创作音乐的能力。<ref name=":24" />IMEC发布了由芯片原型机谱写的30秒旋律。向芯片加载一系列特征、风格统一的歌曲（古代比利时和法国长笛小步舞曲），芯片从中学习相关规则并将其应用于创作。<ref name=":25" /><br />
<br />
[[Blue Brain Project|The Blue Brain Project]], led by Henry Markram, aims to build biologically detailed digital reconstructions and simulations of the mouse brain. The Blue Brain Project has created in silico models of rodent brains, while attempting to replicate as many details about its biology as possible. The supercomputer-based simulations offer new perspectives on understanding the structure and functions of the brain.<br />
<br />
由亨利·马克拉姆领导的'''<font color="#ff8000">蓝脑计划The Blue Brain Project</font>'''旨在建立小鼠大脑生理学细节的数字重建和模拟。蓝脑计划已经建立了啮齿动物大脑的电子模型，同时进行着尽可能多地复制其生理学细节的尝试。基于超级计算机的模拟为理解大脑的结构和功能提供了新的视角。<br />
<br />
The European Union funded a series of projects at the University of Heidelberg, which led to the development of <br />
[[BrainScaleS]] (brain-inspired multiscale computation in neuromorphic hybrid systems), a hybrid analog [[neuromorphic]] supercomputer located at Heidelberg University, Germany. It was developed as part of the [[Human Brain Project]] neuromorphic computing platform and is the complement to the [[SpiNNaker]] supercomputer (which is based on digital technology). The architecture used in BrainScaleS mimics biological neurons and their connections on a physical level; additionally, since the components are made of silicon, these model neurons operate on average 864 times (24 hours of real time is 100 seconds in the machine simulation) that of their biological counterparts.<ref name=":26">{{Cite web|date=2016-03-21|title=Beyond von Neumann, Neuromorphic Computing Steadily Advances|url=https://www.hpcwire.com/2016/03/21/lacking-breakthrough-neuromorphic-computing-steadily-advance/|access-date=2021-10-08|website=HPCwire|language=en-US}}</ref><br />
<br />
欧盟资助了海德堡大学一系列促进BrainScaleS（神经形态混合系统中受大脑启发的多尺度计算）发展的项目，这是一台位于德国海德堡大学的混合模拟'''<font color="#ff8000">神经形态Neuromorphic</font>'''超级计算机。它是作为人类大脑计划中神经形态计算平台的一部分而开发的，是'''<font color="#ff8000">SpiNNaker</font>'''超级计算机（基于数字技术）的补充。BrainScaleS中使用的体系架构模拟了生物神经元及其在物理层面上的连接；此外，由于这些组件是由硅制成的，这些模型神经元平均运行速度是生物神经元的864倍，这意味着在机器模拟中，24小时的实时时间仅为100秒。<ref name=":26" /><br />
<br />
===Neuromorphic sensors ===<br />
===神经形态传感器===<br />
The concept of neuromorphic systems can be extended to sensors (not just to computation). An example of this applied to detecting [[light]] is the [[retinomorphic sensor]] or, when employed in an array, the [[event camera]].<br />
<br />
神经形态系统的概念可以扩展到传感器（而不仅仅是计算单元）。一个用于检测光线的例子是'''<font color="#ff8000">类视网膜传感器Retinomorphic sensor</font>'''，或者'''<font color="#ff8000">事件摄像机Event camera</font>阵列。'''<br />
<br />
<br />
==Ethical considerations==<br />
==伦理问题==<br />
While the interdisciplinary concept of neuromorphic engineering is relatively new, many of the same ethical considerations apply to neuromorphic systems as apply to [[human-like machines]] and [[artificial intelligence]] in general. However, the fact that neuromorphic systems are designed to mimic a [[human brain]] gives rise to unique ethical questions surrounding their usage.<br />
<br />
虽然神经形态工程这个跨学科概念相对较新，但许多适用于'''<font color="#ff8000">类人机器Human-like machines</font>'''和人工智能的伦理讨论在神经形态系统领域也无法避免。另外，神经形态系统是为了模仿人类大脑而设计，这一底层逻辑也导致了一些新的伦理问题。<br />
<br />
However, the practical debate is that neuromorphic hardware as well as artificial "neural networks" are immensely simplified models of how the brain operates or processes information at a much lower [[complex system|complexity in terms of size and functional technology]] and a much more regular structure in terms of [[brain connectivity|connectivity]]. Comparing [[neuromorphic chip]]s to the brain is a very crude comparison similar to comparing a plane to a bird just because they both have wings and a tail. The fact is that neural cognitive systems are many orders of magnitude more [[Energy efficiency (physics)|energy-]] and compute-efficient than current state-of-the-art AI and neuromorphic engineering is an attempt to narrow this gap by inspiring from the brain's mechanism just like many engineering designs have [[bioengineering|bio-inspired features]].<br />
<br />
然而，这种争论的实际情况是，神经形态硬件和人工“神经网络”是对大脑运作或处理信息过程的极其简化的模型，在大小和功能技术方面的'''<font color="#ff8000">复杂性Complexity</font>'''相比而言比较低，在'''<font color="#ff8000">连接性Connectivity</font>'''方面也具有更加规则的结构。将'''<font color="#ff8000">神经形态芯片Neuromorphic chips</font>'''与大脑进行比较是一种非常粗糙的比较，类似于仅仅因为飞机有翅膀和尾巴就将它与鸟进行比较。事实上，当前最先进的人工智能在'''<font color="#ff8000">能耗效率Energy-efficiency</font>'''和计算效率方面距离人脑神经认知系统仍有较大差距，而神经形态工程只是一种通过从大脑机制中获得灵感来缩小这种差距的尝试，就像许多工程设计中都具有'''<font color="#ff8000">生物启发的特征Bio-inspired features</font>'''一样。<br />
<br />
===Democratic concerns===<br />
===公众担忧===<br />
Significant ethical limitations may be placed on neuromorphic engineering due to public perception.<ref name=":27">{{Cite report|url=https://ai100.stanford.edu/sites/g/files/sbiybj9861/f/ai_100_report_0831fnl.pdf|title=Artificial Intelligence and Life in 2030|author=2015 Study Panel|date=September 2016|work=One Hundred Year Study on Artificial Intelligence (AI100)|publisher=Stanford University}}</ref> Special [[Eurobarometer]] 382: Public Attitudes Towards Robots, a survey conducted by the European Commission, found that 60% of [[European Union]] citizens wanted a ban of robots in the care of children, the elderly, or the disabled. Furthermore, 34% were in favor of a ban on robots in education, 27% in healthcare, and 20% in leisure. The European Commission classifies these areas as notably “human.” The report cites increased public concern with robots that are able to mimic or replicate human functions. Neuromorphic engineering, by definition, is designed to replicate the function of the human brain.<ref name=":1">{{Cite web|url=http://ec.europa.eu/commfrontoffice/publicopinion/archives/ebs/ebs_382_en.pdf|title=Special Eurobarometer 382: Public Attitudes Towards Robots|last=European Commission|date=September 2012|website=European Commission}}</ref><br />
<br />
由于公众认知的相关忧虑，神经形态工程学可能会受到严重的伦理限制。<ref name=":27" />欧盟委员会进行的一项调查发现，60% 的欧盟公民希望禁止机器人参与照顾儿童、老人或残疾人的工作。此外，34% 的人支持禁止机器人用于教育，27% 的人支持禁止机器人用于医疗保健，20% 的人支持禁止机器人用于娱乐。欧盟委员会将以上领域划入“人类”范畴。报告指出，公众越来越关注能够模仿或复制人类行为的机器人。而神经形态工程，顾名思义，是为了模仿人脑的功能而设计的。<ref name=":1" /><br />
<br />
The democratic concerns surrounding neuromorphic engineering are likely to become even more profound in the future. The European Commission found that EU citizens between the ages of 15 and 24 are more likely to think of robots as human-like (as opposed to instrument-like) than EU citizens over the age of 55. When presented an image of a robot that had been defined as human-like, 75% of EU citizens aged 15–24 said it corresponded with the idea they had of robots while only 57% of EU citizens over the age of 55 responded the same way. The human-like nature of neuromorphic systems, therefore, could place them in the categories of robots many EU citizens would like to see banned in the future.<ref name=":1" /><br />
<br />
围绕神经形态工程的公众担忧可能在未来变得更加严重。欧盟委员会发现，相比于55岁以上的欧盟公民，15至24岁的欧盟公民更有可能认为机器人像人（而不是像仪器）。当看到一张“类人”机器人的图片时，年龄在15岁至24岁之间的欧盟公民中有75% 的人表示这符合他们对机器人的想法，而55岁以上的欧盟公民中只有57% 的人有同样的反应。因此，神经形态系统可能因为其类似人类的特性而被归入许多欧盟公民希望在未来禁止使用的机器人类别。<ref name=":1" /><br />
<br />
===Personhood===<br />
===人格权问题===<br />
As neuromorphic systems have become increasingly advanced, some scholars{{who|date=August 2021}} have advocated for granting [[personhood]] rights to these systems. If the brain is what grants humans their personhood, to what extent does a neuromorphic system have to mimic the human brain to be granted personhood rights? Critics of technology development in the [[Human Brain Project]], which aims to advance brain-inspired computing, have argued that advancement in neuromorphic computing could lead to machine consciousness or personhood.<ref name=":28">{{Cite journal|last=Aicardi|first=Christine|date=September 2018|title=Accompanying technology development in the Human Brain Project: From foresight to ethics management|journal=Futures|volume=102|pages=114–124|doi=10.1016/j.futures.2018.01.005|doi-access=free}}</ref> If these systems are to be treated as people, critics argue, then many tasks humans perform using neuromorphic systems, including the act of termination of neuromorphic systems, may be morally impermissible as these acts would violate the autonomy of the neuromorphic systems.<ref name=":29">{{Cite journal|last=Lim|first=Daniel|date=2014-06-01|title=Brain simulation and personhood: a concern with the Human Brain Project|journal=Ethics and Information Technology|language=en|volume=16|issue=2|pages=77–89|doi=10.1007/s10676-013-9330-5|s2cid=17415814|issn=1572-8439}}</ref><br />
<br />
随着神经形态系统的日益发展，一些学者主张赋予这些系统'''<font color="ff8000">人格Personhood</font>'''权。如果是大脑赋予了人类人格，那么在多大程度上模仿人类大脑的神经形态系统才能被赋予人格权利？“人类大脑计划”旨在推进以大脑为灵感的计算机技术发展，该计划的批评者认为，神经形态计算机技术的进步可能导致机器意识或人格的形成。<ref name=":28" />这些批评者认为，如果这些系统被当作人来对待，那么人类使用神经形态系统执行任务（包括终止神经形态系统）的行为，在道德上就可能是不被允许的，因为这些行为将违反神经形态系统的自主性。<ref name=":29" /><br />
<br />
<br />
==Dual use (military applications)==<br />
==军民两用技术（军事应用）==<br />
The [[Joint Artificial Intelligence Center]], a branch of the U.S. military, is a center dedicated to the procurement and implementation of AI software and neuromorphic hardware for combat use. Specific applications include smart headsets/goggles and robots. JAIC intends to rely heavily on neuromorphic technology to connect "every fighter every shooter" within a network of neuromorphic-enabled units.<br />
<br />
'''<font color="#ff8000">联合人工智能中心The Joint Artificial Intelligence Center</font>'''（JAIC），是美国军队的一个分支，专门从事采购和实施用于战斗的人工智能软件和神经形态硬件。具体应用包括智能耳机、护目镜和机器人。JAIC打算高度依赖神经形态技术，'''<font color="#32CD32">使用神经形态技术来连接神经形态单位网络中的“每个战士每个射手”</font>'''。<br />
<br />
<br />
==Legal considerations==<br />
==法律问题==<br />
Skeptics have argued that there is no way to apply the electronic personhood, the concept of personhood that would apply to neuromorphic technology, legally. In a letter signed by 285 experts in law, robotics, medicine, and ethics opposing a European Commission proposal to recognize “smart robots” as legal persons, the authors write, “A legal status for a robot can’t derive from the [[Natural person|Natural Person]] model, since the robot would then hold [[human rights]], such as the right to dignity, the right to its integrity, the right to remuneration or the right to citizenship, thus directly confronting the Human rights. This would be in contradiction with the [[Charter of Fundamental Rights of the European Union]] and the [[Convention for the Protection of Human Rights and Fundamental Freedoms]].”<ref name=":30">{{Cite web|url=http://www.robotics-openletter.eu/|title=Robotics Openletter {{!}} Open letter to the European Commission|language=fr-FR|access-date=2019-05-10}}</ref><br />
<br />
怀疑派认为，在法律上没有办法应用能够适用于神经形态技术的电子人格。在一封由285名法律、机器人技术、医学和伦理学专家的联名信中，作者们反对欧盟委员会提出的承认“智能机器人”为法人的提议。他们写道，“机器人的法律地位无法从'''<font color="#ff8000">自然人Natural Person</font>'''模型中推导出来，因为机器人将被赋予'''<font color="#ff8000">人权Human rights</font>'''，如尊严权、完整权、报酬权或公民权，从而直接面临人权问题。这将有悖于《欧联基本权利宪章》和《欧洲保障人权和根本自由公约》”。<ref name=":30" /><br />
<br />
===Ownership and property rights ===<br />
===所有权及财产权问题===<br />
There is significant legal debate around property rights and artificial intelligence. In ''Acohs Pty Ltd v. Ucorp Pty Ltd'', Justice Christopher Jessup of the [[Federal Court of Australia]] found that the [[source code]] for [[Material safety data sheets|Material Safety Data Sheets]] could not be [[Copyright law of Australia|copyrighted]] as it was generated by a [[software interface]] rather than a human author.<ref name=":31">{{Cite web|url=http://www.lavan.com.au/advice/intellectual_property_technology/copyright_in_source_code_and_digital_products|title=Copyright in source code and digital products|last=Lavan|website=Lavan|language=en|access-date=2019-05-10}}</ref> The same question may apply to neuromorphic systems: if a neuromorphic system successfully mimics a human brain and produces a piece of original work, who, if anyone, should be able to claim ownership of the work?<ref name=":32">{{cite journal |last1=Eshraghian|first1=Jason K. |title=Human Ownership of Artificial Creativity |journal=Nature Machine Intelligence |date=9 March 2020 |volume=2 |pages=157–160 |doi=10.1038/s42256-020-0161-x}}</ref><br />
<br />
法律界围绕财产权和人工智能有着重大争论。在Acohs Pty有限公司诉Ucorp Pty有限公司一案中，澳大利亚联邦法院的克里斯托弗·杰瑟普法官发现，版权保护不适用于材料安全数据表的'''<font color="#ff8000">源代码Source code</font>'''，因为它是由'''<font color="#ff8000">软件界面Software interface</font>'''生成而非人类工作者生成的。<ref name=":31" />同样的问题可能也适用于神经形态系统：如果一个神经形态系统成功地模仿了人类的大脑，并产生了一部原创作品，那么该如何确认这部作品的所有权归属？<ref name=":32" /><br />
<br />
<br />
==Neuromemristive systems==<br />
==神经忆阻系统==<br />
Neuromemristive systems are a subclass of neuromorphic computing systems that focus on the use of [[memristors]] to implement [[neuroplasticity]]. While neuromorphic engineering focuses on mimicking biological behavior, neuromemristive systems focus on abstraction.<ref name=":33">{{Cite web|url=https://digitalops.sandia.gov/Mediasite/Play/a10cf6ceb55d47608bb8326dd00e46611d|title=002.08 N.I.C.E. Workshop 2014: Towards Intelligent Computing with Neuromemristive Circuits and Systems - Feb. 2014|website=digitalops.sandia.gov|access-date=2019-08-26}}</ref> For example, a neuromemristive system may replace the details of a [[Cerebral cortex|cortical]] microcircuit's behavior with an abstract neural network model.<ref name=":34">C. Merkel and D. Kudithipudi, "Neuromemristive extreme learning machines for pattern classification," ISVLSI, 2014.</ref><br />
<br />
神经忆阻系统是神经形态计算系统的一个亚类，主要研究利用'''<font color="ff8000">忆阻器Memristors</font>'''实现'''<font color="ff8000">神经可塑性Neuroplasticity</font>'''。神经形态工程的重点是模拟生物行为，而神经忆阻系统的重点是提取。<ref name=":33" />举个例子，一个神经忆阻系统可能用抽象的神经网络模型替代'''<font color="ff8000">皮层Cortical</font>'''微电路的行为细节。<ref name=":34" /><br />
<br />
There exist several neuron inspired threshold logic functions<ref name="Maan 1–13" /> implemented with memristors that have applications in high level [[pattern recognition]] applications. Some of the applications reported recently include [[speech recognition]],<ref name=":35">{{Cite journal|title = Memristor pattern recogniser: isolated speech word recognition|journal = Electronics Letters|pages = 1370–1372|volume = 51|issue = 17|doi = 10.1049/el.2015.1428|first1 = A.K.|last1 = Maan|first2 = A.P.|last2 = James|first3 = S.|last3 = Dimitrijev|year = 2015|bibcode = 2015ElL....51.1370M|hdl = 10072/140989|s2cid = 61454815|url = https://semanticscholar.org/paper/48d3ab11ec6e213b62f11eedcfb7b7febb058674|hdl-access = free}}</ref> [[face recognition]]<ref name=":36">{{Cite journal|title = Memristive Threshold Logic Face Recognition|journal = Procedia Computer Science|date = 2014-01-01|pages = 98–103|volume = 41|series = 5th Annual International Conference on Biologically Inspired Cognitive Architectures, 2014 BICA|doi = 10.1016/j.procs.2014.11.090|first1 = Akshay Kumar|last1 = Maan|first2 = Dinesh S.|last2 = Kumar|first3 = Alex Pappachen|last3 = James|doi-access = free}}</ref> and [[object recognition]].<ref name=":37">{{Cite journal|title = Memristive Threshold Logic Circuit Design of Fast Moving Object Detection|journal = IEEE Transactions on Very Large Scale Integration (VLSI) Systems|date = 2015-10-01|issn = 1063-8210|pages = 2337–2341|volume = 23|issue = 10|doi = 10.1109/TVLSI.2014.2359801|first1 = A.K.|last1 = Maan|first2 = D.S.|last2 = Kumar|first3 = S.|last3 = Sugathan|first4 = A.P.|last4 = James|arxiv = 1410.1267|s2cid = 9647290}}</ref> They also find applications in replacing conventional digital logic gates.<ref name=":38">{{Cite journal|title = Resistive Threshold Logic|journal = IEEE Transactions on Very Large Scale Integration (VLSI) Systems|date = 2014-01-01|issn = 1063-8210|pages = 190–195|volume = 22|issue = 1|doi = 10.1109/TVLSI.2012.2232946|first1 = A.P.|last1 = James|first2 = L.R.V.J.|last2 = Francis|first3 = D.S.|last3 = Kumar|arxiv = 1308.0090|s2cid = 7357110}}</ref><ref name=":39">{{Cite journal|title = Threshold Logic Computing: Memristive-CMOS Circuits for Fast Fourier Transform and Vedic Multiplication|journal = IEEE Transactions on Very Large Scale Integration (VLSI) Systems|date = 2015-11-01|issn = 1063-8210|pages = 2690–2694|volume = 23|issue = 11|doi = 10.1109/TVLSI.2014.2371857|first1 = A.P.|last1 = James|first2 = D.S.|last2 = Kumar|first3 = A.|last3 = Ajayan|arxiv = 1411.5255|s2cid = 6076956}}</ref><br />
<br />
受神经元启发、使用忆阻器实现的阈值逻辑函数<ref name="Maan 1–13" />在高级'''<font color="#ff8000">模式识别Pattern recognition</font>'''中有着广泛的应用，最近报道中其应用包括'''<font color="#ff8000">语音识别Speech recognition<ref name=":35" /></font>'''、'''<font color="#ff8000">人脸识别Face recognition<ref name=":36" /></font>'''和'''<font color="#ff8000">物体识别Object recognition<ref name=":37" /></font>'''。阈值逻辑函数还可以用来取代传统的数字逻辑门。<ref name=":38" /><ref name=":39" /><br />
<br />
For ideal passive memristive circuits there is an exact equation (Caravelli-Traversa-[[Di Ventra]] equation) for the internal memory of the circuit:<ref name=":40">{{cite journal |last=Caravelli |display-authors=etal|arxiv=1608.08651 |title=The complex dynamics of memristive circuits: analytical results and universal slow relaxation |year=2017 |doi=10.1103/PhysRevE.95.022140 |pmid= 28297937 |volume=95 |issue= 2 |pages= 022140 |journal=Physical Review E|bibcode=2017PhRvE..95b2140C |s2cid=6758362}}</ref><br />
<br />
对于理想的无源记忆电路，电路的内部记忆可以用精确的方程（Caravelli-Traversa-Di Ventra方程) 来描述:<ref name=":40" /><br />
<br />
:<math> \frac{d}{dt} \vec{W} = \alpha \vec{W}-\frac{1}{\beta} (I+\xi \Omega W)^{-1} \Omega \vec S </math><br />
<br />
as a function of the properties of the physical memristive network and the external sources. In the equation above, <math>\alpha</math> is the "forgetting" time scale constant, <math> \xi=r-1</math> and <math>r=\frac{R_\text{off}}{R_\text{on}}</math> is the ratio of ''off'' and ''on'' values of the limit resistances of the memristors, <math> \vec S </math> is the vector of the sources of the circuit and <math>\Omega</math> is a projector on the fundamental loops of the circuit. The constant <math>\beta</math> has the dimension of a voltage and is associated to the properties of the [[memristor]]; its physical origin is the charge mobility in the conductor. The diagonal matrix and vector <math>W=\operatorname{diag}(\vec W)</math> and <math>\vec W</math> respectively, are instead the internal value of the memristors, with values between 0 and 1. This equation thus requires adding extra constraints on the memory values in order to be reliable.<br />
<br />
Caravelli-Traversa-Di Ventra方程是描述物理记忆网络和外部源性质的函数。在上述方程中，<math>\alpha</math>是“遗忘”时间尺度常数，<math>\xi=r-1</math>，<math>r =\frac{R\text_{off}}{R_\text{on}}</math>是记忆电阻器off状态和on状态极限电阻值之比，<math>\vec S</math>是电路源的矢量，<math>\Omega</math>是电路基本环路的投影。常数<math>\beta</math>具有电压的量纲，与记忆电阻器的特性有关；其物理原型是导体中的电荷迁移率。对角矩阵和向量 <math>W=\operatorname{diag}(\vec W)</math>和<math>\vec W</math> '''<font color="#32CD32">分别是忆阻器的内阻</font>'''，值在0到1之间。因此，这个等式需要在'''<font color="32CD32">内存值</font>'''上添加额外约束以保证可靠性。<br />
<br />
<br />
==See also==<br />
==相关词条==<br />
{{Columns-list|colwidth=18em|<br />
* [[AI accelerator (computer hardware)|AI accelerator]]<br />
* [[Artificial brain]]<br />
* [[Biomorphic]]<br />
* [[Cognitive computer]]<br />
* [[Computation and Neural Systems]]<br />
* [[Differentiable programming]]<br />
* [[Event camera]]<br />
* [[Neurorobotics]]<br />
* [[Optical flow sensor]]<br />
* [[Physical neural network]]<br />
* [[SpiNNaker]]<br />
* [[SyNAPSE]]<br />
* [[Retinomorphic sensor]]<br />
* [[Vision chip]]<br />
* [[Vision processing unit]]<br />
* [[Zeroth (software)]]<br />
* [[Hardware for artificial intelligence]]<br />
}}<br />
<br />
{{Portal bar|Electronics}}<br />
<br />
==References==<br />
{{Reflist|40em}}<br />
<br />
== External links ==<br />
<!--======================== {{No more links}} ============================<br />
| PLEASE BE CAUTIOUS IN ADDING MORE LINKS TO THIS ARTICLE. Wikipedia |<br />
| is not a collection of links nor should it be used for advertising. |<br />
| |<br />
| Excessive or inappropriate links WILL BE DELETED. |<br />
| See [[Wikipedia:External links]] & [[Wikipedia:Spam]] for details. |<br />
| |<br />
| If there are already plentiful links, please propose additions or |<br />
| replacements on this article's discussion page, or submit your link |<br />
| to the relevant category at the Open Directory Project (dmoz.org) |<br />
| and link back to that category using the {{dmoz}} template. |<br />
======================={{No more links}}=============================--><br />
{{Commons category|Neuromorphic Engineering}}<br />
*[https://web.archive.org/web/20150727034331/http://ine-web.org/workshops/workshops-overview Telluride Neuromorphic Engineering Workshop]<br />
*[https://archive.today/20130115190057/http://capocaccia.ethz.ch/ CapoCaccia Cognitive Neuromorphic Engineering Workshop]<br />
*[https://web.archive.org/web/20190716132350/http://www.ine-web.org/ Institute of Neuromorphic Engineering]<br />
*[http://www.ine-news.org/ INE news site].<br />
*[http://www.frontiersin.org/neuromorphic_engineering Frontiers in Neuromorphic Engineering Journal]<br />
*[http://www.cns.caltech.edu/ Computation and Neural Systems] department at the [[California Institute of Technology]].<br />
*[http://www.humanbrainproject.eu/ Human Brain Project official site]<br />
*[https://www.the-scientist.com/features/building-a-silicon-brain-65738 Building a Silicon Brain:] Computer chips based on biological neurons may help simulate larger and more-complex brain models. May 1, 2019. SANDEEP RAVINDRAN<br />
<br />
<br />
*Telluride Neuromorphic Engineering Workshop<br />
*CapoCaccia Cognitive Neuromorphic Engineering Workshop<br />
*Institute of Neuromorphic Engineering<br />
*INE news site.<br />
* Frontiers in Neuromorphic Engineering Journal<br />
*Computation and Neural Systems department at the California Institute of Technology.<br />
* Human Brain Project official site<br />
*Building a Silicon Brain: Computer chips based on biological neurons may help simulate larger and more-complex brain models. May 1, 2019. SANDEEP RAVINDRAN<br />
<br />
=<nowiki>外部链接</nowiki>= <br />
*碲化物神经形态工程工作室<br />
*CapoCaccia认知神经形态工程工作室<br />
*神经形态工程研究所<br />
*INE新闻站点。<br />
*《神经形态工程学前沿》<br />
*加州理工学院计算与神经系统系。<br />
*人脑项目官方网站<br />
*打造硅脑: 基于生物神经元的计算机芯片可能有助于模拟更大、更复杂的大脑模型。2019年5月1日。SANDEEP RAVINDRAN.<br />
<br />
{{Differentiable computing}}<br />
{{Authority control}}<br />
<br />
[[Category:Electrical engineering]]<br />
[[Category:Neuroscience]]<br />
[[分类:AI accelerators|*]]<br />
[[Category:Artificial intelligence]]<br />
[[Category:Robotics]]<br />
<br />
Category:Electrical engineering<br />
Category:Neuroscience<br />
*<br />
Category:Artificial intelligence<br />
Category:Robotics<br />
<br />
类别: 电气工程类别: 神经科学 * 类别: 人工智能类别: 机器人<br />
<br />
<noinclude><br />
<br />
<small>This page was moved from [[wikipedia:en:Neuromorphic engineering]]. Its edit history can be viewed at [[神经形态计算/edithistory]]</small></noinclude><br />
<br />
[[Category:待整理页面]]<br />
<br />
<references /></div>Shenky20https://wiki.swarma.org/index.php?title=%E7%A5%9E%E7%BB%8F%E5%BD%A2%E6%80%81%E8%AE%A1%E7%AE%97&diff=30754神经形态计算2022-04-28T12:00:29Z<p>Shenky20：/* 神经记忆电阻系统 */</p>
<hr />
<div><br />
此词条由神经动力学模型读书会词条梳理志愿者Shenky20翻译审校，未经专家审校，带来阅读不便，请见谅。<br />
<br />
{{Use American English|date = January 2019}}<br />
{{Use mdy dates|date = January 2019}}<br />
<br />
'''Neuromorphic engineering''', also known as '''neuromorphic computing''',<ref name=":3">{{Cite journal | doi = 10.1145/2601069| title = Neuromorphic computing gets ready for the (really) big time| journal = [[Communications of the ACM]]| volume = 57| issue = 6| year = 2014| last1 = Monroe | first1 = D. | pages = 13–15| s2cid = 20051102}}</ref><ref name=":4">{{Cite journal | doi = 10.1088/0957-4484/21/17/175202| pmid = 20368686| title = Nanotube devices based crossbar architecture: Toward neuromorphic computing| journal = Nanotechnology| volume = 21| issue = 17| pages = 175202| year = 2010| last1 = Zhao | first1 = W. S. | last2 = Agnus | first2 = G. | last3 = Derycke | first3 = V. | last4 = Filoramo | first4 = A. | last5 = Bourgoin | first5 = J. -P. | last6 = Gamrat | first6 = C. | bibcode = 2010Nanot..21q5202Z| url = https://zenodo.org/record/3428659}}</ref><ref name="humanbrainproject">{{YouTube|id=6RoiZ90mGfw|title=The Human Brain Project SP 9: Neuromorphic Computing Platform}}</ref> is the use of [[very-large-scale integration]] (VLSI) systems containing electronic [[analog circuit]]s to mimic neuro-biological architectures present in the nervous system. A neuromorphic computer/chip is any device that uses physical artificial neurons (made from silicon) to do computations.<ref name=":5">{{cite journal|last1=Mead|first1=Carver|title=Neuromorphic electronic systems|journal=Proceedings of the IEEE|date=1990|volume=78|issue=10|pages=1629–1636|doi=10.1109/5.58356|url=https://authors.library.caltech.edu/53090/1/00058356.pdf}}</ref><ref name=":2" /> In recent times, the term ''neuromorphic'' has been used to describe analog, digital, mixed-mode analog/digital VLSI, and software systems that implement models of [[neural system]]s (for [[perception]], [[motor control]], or [[multisensory integration]]). The implementation of neuromorphic computing on the hardware level can be realized by oxide-based [[memristor]]s,<ref name="Maan 1–13">{{Cite journal|last1=Maan|first1=A. K.|last2=Jayadevi|first2=D. A.|last3=James|first3=A. P.|date=2016-01-01|title=A Survey of Memristive Threshold Logic Circuits|journal=IEEE Transactions on Neural Networks and Learning Systems|volume=PP|issue=99|pages=1734–1746|doi=10.1109/TNNLS.2016.2547842|pmid=27164608|issn=2162-237X|arxiv=1604.07121|bibcode=2016arXiv160407121M|s2cid=1798273}}</ref> spintronic memories, threshold switches, and [[transistor]]s.<ref name=":6">{{Cite journal|title = Mott Memory and Neuromorphic Devices|journal = Proceedings of the IEEE|date = 2015-08-01|issn = 0018-9219|pages = 1289–1310|volume = 103|issue = 8|doi = 10.1109/JPROC.2015.2431914|first1 = You|last1 = Zhou|first2 = S.|last2 = Ramanathan|s2cid = 11347598|url = https://zenodo.org/record/895565}}</ref><ref name=":2">{{Cite document|title=Neuromorphic Circuits With Neural Modulation Enhancing the Information Content of Neural Signaling {{!}} International Conference on Neuromorphic Systems 2020|language=EN|doi=10.1145/3407197.3407204|s2cid=220794387}}</ref> Training software-based neuromorphic systems of spiking neural networks can be achieved using error backpropagation, e.g., using [[Python (programming language)|Python]] based frameworks such as snnTorch,<ref name=":7">{{cite journal |last1=Eshraghian|first1=Jason K.|last2=Ward|first2=Max|last3=Neftci |first3=Emre|last4=Wang|first4=Xinxin|last5=Lenz|first5=Gregor|last6=Dwivedi|first6=Girish|last7=Bennamoun|first7=Mohammed|last8=Jeong|first8=Doo Seok|last9=Lu|first9=Wei D.|title=Training Spiking Neural Networks Using Lessons from Deep Learning |date=1 October 2021 |arxiv=2109.12894 }}</ref> or using canonical learning rules from the biological learning literature, e.g., using BindsNet.<ref name=":8">{{Cite web | url=https://github.com/Hananel-Hazan/bindsnet | title=Hananel-Hazan/bindsnet: Simulation of spiking neural networks (SNNs) using PyTorch.| date=31 March 2020}}</ref><br />
<br />
<br />
'''<font color="#ff8000">神经形态工程Neuromorphic engineering</font>'''(也称为'''<font color="#ff8000">神经形态计算Neuromorphic computing</font>''')<ref name=":3" /><ref name=":4" /><ref name="humanbrainproject" />是指使用包含电子'''<font color="#ff8000">模拟电路Analog circuit</font>'''的'''<font color="#ff8000">超大规模集成电路Very-large-scale integration</font>'''系统来模拟神经系统中生理结构的研究方法。神经形态计算机或神经形态芯片包括任何使用由硅制成的人造神经元进行计算的设备。<ref name=":5" /><ref name=":2" />近年来，神经形态学（neuromorphic）这个术语被用来描述能够实现'''<font color="#ff8000">神经系统Neural system</font>'''模型功能(如'''<font color="#ff8000">感知Perception</font>'''、'''<font color="#ff8000">运动控制Motor control</font>'''，'''<font color="#ff8000">多感官整合Multisensory integration</font>'''等)的模拟、数字、模拟/数字混合模式超大规模集成电路和软件系统。神经形态计算的硬件实现可以通过基于氧化物的'''<font color="#ff8000">记忆电阻器Memristor</font>'''、<ref name="Maan 1–13" />自旋电子存储器、阈值开关和'''<font color="#ff8000">晶体管Transistor</font>'''来实现。<ref name=":6" /><ref name=":2" />对基于软件的脉冲神经网络系统的训练可以通过误差反向传播机制来实现，例如，使用snnTorch等基于Python的框架，<ref name=":7" />或使用BindsNet等典型的受生物启发的学习模式。<ref name=":8" /><br />
<br />
A key aspect of neuromorphic engineering is understanding how the morphology of individual neurons, circuits, applications, and overall architectures creates desirable computations, affects how information is represented, influences robustness to damage, incorporates learning and development, adapts to local change (plasticity), and facilitates evolutionary change.<br />
<br />
<br />
神经形态工程领域的一个关键问题，就是理解'''<font color="#32CD32">单个神经元形态、神经回路、应用和整体结构</font>'''如何产生理想的计算，如何影响信息的表示和对破坏的鲁棒性，如何整合学习和发展，如何适应局部变化(可塑性) 并促进逐渐发展的变化。<br />
<br />
Neuromorphic engineering is an interdisciplinary subject that takes inspiration from [[biology]], [[physics]], [[mathematics]], [[computer science]], and [[electronic engineering]]<ref name=":2" /> to design artificial neural systems, such as [[Machine vision|vision systems]], head-eye systems, auditory processors, and autonomous robots, whose physical architecture and design principles are based on those of biological nervous systems.<ref name=":9">{{Cite journal | doi = 10.1155/2012/705483| title = Qualitative Functional Decomposition Analysis of Evolved Neuromorphic Flight Controllers| journal = Applied Computational Intelligence and Soft Computing| volume = 2012| pages = 1–21| year = 2012| last1 = Boddhu | first1 = S. K. | last2 = Gallagher | first2 = J. C. | doi-access = free}}</ref> It was developed by [[Carver Mead]]<ref name=":10">{{cite web|title=carver mead website|last1=Mead|first1=Carver|url=http://www.carvermead.caltech.edu/index.html|website=carvermead}}</ref> in the late 1980s.<br />
<br />
<br />
神经形态工程是以'''<font color="#ff8000">生物学Biology</font>'''、'''<font color="#ff8000">物理学Physics</font>'''、'''<font color="#ff8000">数学Mathematics</font>'''、'''<font color="#ff8000">计算机科学Computer science</font>'''和'''<font color="#ff8000">电子工程Electronic engineering<ref name=":2" /></font>'''等学科为基础，设计人工神经系统（如'''<font color="#ff8000">视觉系统Vision systems</font>'''、头眼系统、听觉处理器以及物理结构和设计原则都受启发于生物神经系统的自主机器人）的一门交叉学科。<ref name=":9" />20世纪80年代后期，卡弗·米德极大地推动了神经形态工程领域的发展。<ref name=":10" /><br />
<br />
==Neurological inspiration==<br />
==神经生物学启发==<br />
Neuromorphic engineering is set apart by the inspiration it takes from what we know about the structure and operations of the [[brain]]. Neuromorphic engineering translates what we know about the brain's function into computer systems. Work has mostly focused on replicating the analog nature of [[biological computation]] and the role of [[neuron]]s in [[cognition]].<br />
<br />
<br />
神经形态工程形成的灵感来源于现已知的'''<font color="#ff8000">大脑Brain</font>'''结构及运作机制知识，它将我们对大脑功能的了解用于研究和优化计算机系统。该领域工作主要集中于对'''<font color="#ff8000">生物计算Biological computation</font>'''模拟特性和'''<font color="#ff8000">神经元Neuron</font>'''在'''<font color="#ff8000">认知Cognition</font>'''中所发挥作用的复刻。<br />
<br />
The biological processes of neurons and their [[synapse]]s are dauntingly complex, and thus very difficult to artificially simulate. A key feature of biological brains is that all of the processing in neurons use analog chemical signals. This makes it hard to replicate brains in computers because the current generation of computers is completely digital. However, the characteristics of these parts can be abstracted into mathematical functions that closely capture the essence of the neuron's operations.<br />
<br />
神经元及其'''<font color="#ff8000">突触Synapse</font>'''的生理学过程极其复杂，难以进行人工模拟。神经元在所有处理过程中都使用化学模拟信号，这是大脑的一个关键生理学特征。这个特征大大增加了在计算机中复制大脑的难度，因为目前的计算机是完全数字化的。然而，部分特征可以抽象为数学函数，这些函数能够紧密捕捉神经元运行的本质。<br />
<br />
The goal of neuromorphic computing is not to perfectly mimic the brain and all of its functions, but instead to extract what is known of its structure and operations to be used in a practical computing system. No neuromorphic system will claim nor attempt to reproduce every element of neurons and synapses, but all adhere to the idea that computation is highly [[distributed processing|distributed]] throughout a series of small computing elements analogous to a neuron. While this sentiment is standard, researchers chase this goal with different methods.<ref name=":11">{{Cite journal | doi = 10.1088/1741-2560/13/5/051001| title = Large-scale neuromorphic computing systems| journal = Journal of Neural Engineering| volume = 13| pages = 1–15| year = 2016| last1 = Furber | first1 = Steve| issue = 5| pmid = 27529195| bibcode = 2016JNEng..13e1001F| doi-access = free}}</ref><br />
<br />
神经形态计算的目标不是完美地模拟大脑及其所有功能，而是利用已知的大脑结构和运转机制来研发或优化实际的计算系统。任何神经形态学系统都不会声称或试图复制神经元和突触中的每一个元素，但所有人都一致认可将计算高度'''<font color="#ff8000">分散Distribute</font>'''于一系列类似于神经元的小型计算元素的理念。研究人员用不同的方法来追求这一普遍目标。<ref name=":11" /><br />
<br />
<br />
==Examples==<br />
==代表性成果==<br />
As early as 2006, researchers at [[Georgia Tech]] published a field programmable neural array.<ref name=":12">{{Cite book|title = A field programmable neural array|last1 = Farquhar|first1 = Ethan|date = May 2006|journal = IEEE International Symposium on Circuits and Systems|pages = 4114–4117|last2 = Hasler|first2 = Paul.|doi = 10.1109/ISCAS.2006.1693534|isbn = 978-0-7803-9389-9|s2cid = 206966013}}</ref> This chip was the first in a line of increasingly complex arrays of floating gate transistors that allowed programmability of charge on the gates of [[MOSFET]]s to model the channel-ion characteristics of neurons in the brain and was one of the first cases of a silicon programmable array of neurons.<br />
<br />
早在2006年，佐治亚理工学院的研究人员就研发出了一种现场可编程神经阵列。<ref name=":12" />在此之后，一系列越来越复杂的浮栅晶体管阵列被成功研发出来，这些晶体管阵列可以通过在'''<font color="#ff8000">金属-氧化物半导体效应晶体管MOSFET</font>'''的栅极上编程来模拟大脑中神经元的离子通道特性，这也是以硅为主要材料的可编程神经元阵列的最早成功案例之一。<br />
<br />
In November 2011, a group of [[MIT]] researchers created a computer chip that mimics the analog, ion-based communication in a synapse between two neurons using 400 transistors and standard [[CMOS]] manufacturing techniques.<ref name=":13">{{cite web|title=MIT creates "brain chip"|url=http://www.extremetech.com/extreme/105067-mit-creates-brain-chip|access-date=4 December 2012}}</ref><ref name="Neuromorphic silicon paper">{{cite journal|title=Neuromorphic silicon neurons and large-scale neural networks: challenges and opportunities|doi=10.3389/fnins.2011.00108|pmid=21991244|pmc=3181466|volume=5|pages=108|journal=Frontiers in Neuroscience|year=2011|last1=Poon|first1=Chi-Sang|last2=Zhou|first2=Kuan|doi-access=free}}</ref><br />
<br />
2011年11月，麻省理工学院的一组研究人员研发出一种计算机芯片，该芯片上使用标准的'''<font color="#ff8000">互补金属氧化物半导体CMOS</font>'''制造技术集成了400个晶体管，能够模拟神经元间突触中基于离子的通讯。<ref name=":13" /><ref name="Neuromorphic silicon paper" /><br />
<br />
In June 2012, [[spintronic]] researchers at [[Purdue University]] presented a paper on the design of a neuromorphic chip using [[Spin valve|lateral spin valve]]s and [[memristor]]s. They argue that the architecture works similarly to neurons and can therefore be used to test methods of reproducing the brain's processing. In addition, these chips are significantly more energy-efficient than conventional ones.<ref name="Spin Devices Prop">{{Cite arXiv|title=Proposal For Neuromorphic Hardware Using Spin Devices|eprint=1206.3227|last1=Sharad|first1=Mrigank|last2=Augustine|first2=Charles|last3=Panagopoulos|first3=Georgios|last4=Roy|first4=Kaushik|class=cond-mat.dis-nn|year=2012}}</ref><br />
<br />
2012年6月，普渡大学的'''<font color="#ff8000">自旋电子学Spintronic</font>'''研究人员发表了一篇关于利用'''<font color="#ff8000">侧向自旋阀Lateral spin valves</font>'''和'''<font color="#ff8000">记忆电阻器Memristors</font>'''设计神经形态芯片的论文。他们认为，这种芯片结构的工作原理与神经元相似，因此可以用于大脑运行机制的复刻方法的测试。此外，这些芯片在能耗方面明显优于传统芯片。<ref name="Spin Devices Prop" /><br />
<br />
Research at [[HP Labs]] on Mott memristors has shown that while they can be non-[[Volatile memory|volatile]], the volatile behavior exhibited at temperatures significantly below the [[phase transition]] temperature can be exploited to fabricate a [[neuristor]],<ref name=":0" /> a biologically-inspired device that mimics behavior found in neurons.<ref name=":0">{{Cite journal | doi = 10.1038/nmat3510| pmid = 23241533| title = A scalable neuristor built with Mott memristors| journal = Nature Materials| volume = 12| issue = 2| pages = 114–7| year = 2012| last1 = Pickett | first1 = M. D. | last2 = Medeiros-Ribeiro | first2 = G. | last3 = Williams | first3 = R. S. | bibcode = 2013NatMa..12..114P| s2cid = 16271627| url = https://semanticscholar.org/paper/b6ba6f496ace2b947f111059663e76bb60e9efeb}}</ref> In September 2013, they presented models and simulations that show how the spiking behavior of these neuristors can be used to form the components required for a [[Turing machine]].<ref name=":14">{{cite journal|doi=10.1088/0957-4484/24/38/384002|title=Phase transitions enable computational universality in neuristor-based cellular automata|author1=Matthew D Pickett|author2=R Stanley Williams|name-list-style=amp|date=September 2013|publisher=IOP Publishing Ltd|journal=Nanotechnology|volume=24|issue=38|pmid=23999059|bibcode=2013Nanot..24L4002P|at=384002}}</ref><br />
<br />
'''<font color="#ff8000">惠普实验室HP labs</font>'''在莫特记忆电阻器上的研究表明，尽管它们可以是非'''<font color="#ff8000">易失性Volatile</font>'''的，但是在'''<font color="#ff8000">相变Phase transition</font>'''温度以下时表现出的易失性行为可以被用来制造'''<font color="#ff8000">类神经元电阻器Neuristor</font>'''(一种生物学启发的模仿神经元行为的硬件)<ref name=":0" />。2013年9月，他们通过模型和仿真展示了这些类神经元电阻器的脉冲行为如何产生'''<font color="#ff8000">图灵机Turing machine</font>'''的所需元素。<ref name=":14" /><br />
<br />
[[Neurogrid]], built by ''Brains in Silicon'' at [[Stanford University]],<ref name=":15">{{cite journal|last1=Boahen|first1=Kwabena|title=Neurogrid: A Mixed-Analog-Digital Multichip System for Large-Scale Neural Simulations|journal=Proceedings of the IEEE|date=24 April 2014|volume=102|issue=5|pages=699–716|doi=10.1109/JPROC.2014.2313565|s2cid=17176371}}</ref> is an example of hardware designed using neuromorphic engineering principles. The circuit board is composed of 16 custom-designed chips, referred to as NeuroCores. Each NeuroCore's analog circuitry is designed to emulate neural elements for 65536 neurons, maximizing energy efficiency. The emulated neurons are connected using digital circuitry designed to maximize spiking throughput.<ref name=":16">{{cite journal|doi=10.1038/503022a|pmid = 24201264|title = Neuroelectronics: Smart connections|journal = Nature|volume = 503|issue = 7474|pages = 22–4|year = 2013|last1 = Waldrop|first1 = M. Mitchell|bibcode = 2013Natur.503...22W|doi-access = free}}</ref><ref name=":17">{{cite journal|doi=10.1109/JPROC.2014.2313565|title = Neurogrid: A Mixed-Analog-Digital Multichip System for Large-Scale Neural Simulations|journal = Proceedings of the IEEE|volume = 102|issue = 5|pages = 699–716|year = 2014|last1 = Benjamin|first1 = Ben Varkey|last2 = Peiran Gao|last3 = McQuinn|first3 = Emmett|last4 = Choudhary|first4 = Swadesh|last5 = Chandrasekaran|first5 = Anand R.|last6 = Bussat|first6 = Jean-Marie|last7 = Alvarez-Icaza|first7 = Rodrigo|last8 = Arthur|first8 = John V.|last9 = Merolla|first9 = Paul A.|last10 = Boahen|first10 = Kwabena|s2cid = 17176371}}</ref><br />
<br />
'''<font color="#ff8000">神经栅格Neurogrid<ref name=":15" /></font>'''是由斯坦福大学Brains in Silicon公司研发的、使用神经形态工程原理设计的硬件。该电路板由16个定制设计的芯片组成(NeuroCores)。在设计中，每个NeuroCore芯片的模拟电路对65536个神经元的神经元素进行模拟，以最大限度地提高能量效率。模拟出的神经元通过设计的数字电路连接，以最大化脉冲吞吐量。<ref name=":16" /><ref name=":17" /><br />
<br />
A research project with implications for neuromorphic engineering is the [[Human Brain Project]] that is attempting to simulate a complete human brain in a supercomputer using biological data. It is made up of a group of researchers in neuroscience, medicine, and computing.<ref name=":18">{{cite web|title=Involved Organizations|url=http://www.humanbrainproject.eu/partners.html|access-date=22 February 2013|url-status=dead|archive-url=https://web.archive.org/web/20130302142627/http://www.humanbrainproject.eu/partners.html|archive-date=2 March 2013}}</ref> [[Henry Markram]], the project's co-director, has stated that the project proposes to establish a foundation to explore and understand the brain and its diseases, and to use that knowledge to build new computing technologies. The three primary goals of the project are to better understand how the pieces of the brain fit and work together, to understand how to objectively diagnose and treat brain diseases, and to use the understanding of the human brain to develop neuromorphic computers. That the simulation of a complete human brain will require a supercomputer a thousand times more powerful than today's encourages the current focus on neuromorphic computers.<ref name=":19">{{cite web|title=Human Brain Project|url=http://www.humanbrainproject.eu|access-date=22 February 2013}}</ref> $1.3 billion has been allocated to the project by The [[European Commission]].<ref name=":20">{{cite web|title=The Human Brain Project and Recruiting More Cyberwarriors|url=http://www.marketplace.org/topics/tech/human-brain-project-and-recruiting-more-cyberwarriors|access-date=22 February 2013|date=January 29, 2013}}</ref><br />
<br />
'''<font color="#ff8000">人类大脑计划Human Brain Project</font>'''对神经形态工程具有较大影响，其主要任务是尝试用生物数据在超级计算机中模拟完整的人脑。人类大脑计划由神经科学、医学和计算机科学背景的研究人员组成。<ref name=":18" />该项目的联合主管亨利•马克拉姆(Henry Markram)表示，人类大脑计划的目的是建立一个探索和了解脑科学和脑疾病知识的基础，并利用这些知识来构建更先进的计算机技术。这个项目的三个主要目标分别是: 更好地理解大脑的各个部分是如何相互配合协同工作的; 理解如何客观地诊断和治疗脑部疾病; 以及利用对人类大脑的理解来开发神经形态计算机。模拟一个完整的人类大脑需要一台比现在强大一千倍的超级计算机，这不断激发着对神经形态计算机领域的研究兴趣。<ref name=":19" />欧盟委员会已经向人类大脑计划拨款13亿美元。<ref name=":20" /><br />
<br />
Other research with implications for neuromorphic engineering involves the [[BRAIN Initiative]]<ref name="economist">[https://www.economist.com/news/science-and-technology/21582495-computers-will-help-people-understand-brains-better-and-understanding-brains Neuromorphic computing: The machine of a new soul], The Economist, 2013-08-03</ref> and the [[TrueNorth]] chip from [[IBM]].<ref name=":21">{{cite journal|last1=Modha|first1=Dharmendra|title=A million spiking-neuron integrated circuit with a scalable communication network and interface|journal=Science|date=Aug 2014|volume=345|issue=6197|pages=668–673|doi=10.1126/science.1254642|pmid=25104385|bibcode=2014Sci...345..668M|s2cid=12706847}}</ref> Neuromorphic devices have also been demonstrated using nanocrystals, nanowires, and conducting polymers.<ref name=":22">{{Cite web|url=http://jessamynfairfield.com/wp-content/uploads/2017/03/PWMar17Fairfield.pdf|title=Smarter Machines|last=Fairfield|first=Jessamyn|date=March 1, 2017}}</ref><br />
<br />
其他与神经形态工程有关的研究还包括'''<font color="#ff8000">脑计划BRAIN initiative</font>'''，<ref name="economist" />和IBM研发的'''<font color="#ff8000">TrueNorth</font>'''芯片。<ref name=":21" />使用纳米晶体、纳米线和导电聚合物也能够用于制造神经形态学硬件。<ref name=":22" /><br />
<br />
[[Intel]] unveiled its neuromorphic research chip, called “[[Intel Loihi|Loihi]]”, in October 2017. The chip uses an asynchronous [[spiking neural network]] (SNN) to implement adaptive self-modifying event-driven fine-grained parallel computations used to implement learning and inference with high efficiency.<ref name=":23">{{cite journal |last1=Davies |first1=Mike |title=Loihi: A Neuromorphic Manycore Processor with On-Chip Learning |journal=IEEE Micro |date=January 16, 2018 |volume=38 |issue=1 |pages=82–99 |display-authors=etal|doi=10.1109/MM.2018.112130359 |s2cid=3608458 }}</ref><ref name="Morris2017">{{cite web |last1=Morris |first1=John |title=Why Intel built a neuromorphic chip |url=https://www.zdnet.com/article/why-intel-built-a-neuromorphic-chip/ |website=ZDNet |access-date=17 August 2018 |language=en}}</ref><br />
<br />
2017年10月，英特尔发布了神经形态芯片'''<font color="#ff8000">Loihi</font>'''。该芯片采用异步'''<font color="#ff8000">脉冲神经网络Spiking neural network</font>'''实现了自适应、自修改、事件驱动的细粒度并行计算，实现了高效的学习和推理。<ref name=":23" /><ref name="Morris2017" /><br />
<br />
[[IMEC]], a Belgium-based nanoelectronics research center, demonstrated the world's first self-learning neuromorphic chip. The brain-inspired chip, based on OxRAM technology, has the capability of self-learning and has been demonstrated to have the ability to compose music.<ref name=":24">{{cite web |title=Imec demonstrates self-learning neuromorphic chip that composes music |url=https://www.imec-int.com/en/articles/imec-demonstrates-self-learning-neuromorphic-chip-that-composes-music |website=IMEC International |access-date=1 October 2019}}</ref> IMEC released the 30-second tune composed by the prototype. The chip was sequentially loaded with songs in the same time signature and style. The songs were old Belgian and French flute minuets, from which the chip learned the rules at play and then applied them.<ref name=":25">{{cite web|last1=Bourzac|first1=Katherine|title=A Neuromorphic Chip That Makes Music|url=https://spectrum.ieee.org/a-neuromorphic-chip-that-makes-music|url-status=live|access-date=1 October 2019|website=IEEE Spectrum|date=May 23, 2017}}</ref><br />
<br />
比利时的'''<font color="#ff8000">微电子研究中心IMEC</font>'''研发了世界上首个自学习神经形态芯片。这种基于'''<font color="#32CD32"> OxRAM(filamentary-oxide-based resistive memory technology)</font>''' 技术的大脑启发芯片具有自学习能力，并且已被证明具有创作音乐的能力。<ref name=":24" />IMEC发布了由芯片原型机谱写的30秒旋律。向芯片加载一系列特征、风格统一的歌曲（古代比利时和法国长笛小步舞曲），芯片从中学习相关规则并将其应用于创作。<ref name=":25" /><br />
<br />
[[Blue Brain Project|The Blue Brain Project]], led by Henry Markram, aims to build biologically detailed digital reconstructions and simulations of the mouse brain. The Blue Brain Project has created in silico models of rodent brains, while attempting to replicate as many details about its biology as possible. The supercomputer-based simulations offer new perspectives on understanding the structure and functions of the brain.<br />
<br />
由亨利·马克拉姆领导的'''<font color="#ff8000">蓝脑计划The Blue Brain Project</font>'''旨在建立小鼠大脑生理学细节的数字重建和模拟。蓝脑计划已经建立了啮齿动物大脑的电子模型，同时进行着尽可能多地复制其生理学细节的尝试。基于超级计算机的模拟为理解大脑的结构和功能提供了新的视角。<br />
<br />
The European Union funded a series of projects at the University of Heidelberg, which led to the development of <br />
[[BrainScaleS]] (brain-inspired multiscale computation in neuromorphic hybrid systems), a hybrid analog [[neuromorphic]] supercomputer located at Heidelberg University, Germany. It was developed as part of the [[Human Brain Project]] neuromorphic computing platform and is the complement to the [[SpiNNaker]] supercomputer (which is based on digital technology). The architecture used in BrainScaleS mimics biological neurons and their connections on a physical level; additionally, since the components are made of silicon, these model neurons operate on average 864 times (24 hours of real time is 100 seconds in the machine simulation) that of their biological counterparts.<ref name=":26">{{Cite web|date=2016-03-21|title=Beyond von Neumann, Neuromorphic Computing Steadily Advances|url=https://www.hpcwire.com/2016/03/21/lacking-breakthrough-neuromorphic-computing-steadily-advance/|access-date=2021-10-08|website=HPCwire|language=en-US}}</ref><br />
<br />
欧盟资助了海德堡大学一系列促进BrainScaleS(神经形态混合系统中受大脑启发的多尺度计算)发展的项目，这是一台位于德国海德堡大学的混合模拟'''<font color="#ff8000">神经形态Neuromorphic</font>'''超级计算机。它是作为人类大脑计划中神经形态计算平台的一部分而开发的，是'''<font color="#ff8000">SpiNNaker</font>'''超级计算机(基于数字技术)的补充。BrainScaleS中使用的体系架构模拟了生物神经元及其在物理层面上的连接；此外，由于这些组件是由硅制成的，这些模型神经元平均运行速度是生物神经元的864倍，这意味着在机器模拟中，24小时的实时时间仅为100秒。<ref name=":26" /><br />
<br />
===Neuromorphic sensors===<br />
===神经形态传感器===<br />
The concept of neuromorphic systems can be extended to sensors (not just to computation). An example of this applied to detecting [[light]] is the [[retinomorphic sensor]] or, when employed in an array, the [[event camera]].<br />
<br />
神经形态系统的概念可以扩展到传感器(而不仅仅是计算单元)。一个用于检测光线的例子是'''<font color="#ff8000">类视网膜传感器Retinomorphic sensor</font>'''，或者'''<font color="#ff8000">事件摄像机Event camera</font>阵列。'''<br />
<br />
<br />
==Ethical considerations ==<br />
==伦理问题==<br />
While the interdisciplinary concept of neuromorphic engineering is relatively new, many of the same ethical considerations apply to neuromorphic systems as apply to [[human-like machines]] and [[artificial intelligence]] in general. However, the fact that neuromorphic systems are designed to mimic a [[human brain]] gives rise to unique ethical questions surrounding their usage.<br />
<br />
虽然神经形态工程这个跨学科概念相对较新，但许多适用于'''<font color="#ff8000">类人机器Human-like machines</font>'''和人工智能的伦理考虑大体上也适用于神经形态系统。另一方面，神经形态系统是为了模仿人类大脑而设计，这一事实引起了一些针对性的独特伦理问题。<br />
<br />
However, the practical debate is that neuromorphic hardware as well as artificial "neural networks" are immensely simplified models of how the brain operates or processes information at a much lower [[complex system|complexity in terms of size and functional technology]] and a much more regular structure in terms of [[brain connectivity|connectivity]]. Comparing [[neuromorphic chip]]s to the brain is a very crude comparison similar to comparing a plane to a bird just because they both have wings and a tail. The fact is that neural cognitive systems are many orders of magnitude more [[Energy efficiency (physics)|energy-]] and compute-efficient than current state-of-the-art AI and neuromorphic engineering is an attempt to narrow this gap by inspiring from the brain's mechanism just like many engineering designs have [[bioengineering|bio-inspired features]].<br />
<br />
然而，这种争论的实际情况是，神经形态硬件和人工“神经网络”是对大脑运作或处理信息过程的极其简化的模型，在大小和功能技术方面的'''<font color="#ff8000">复杂性Complexity</font>'''相比而言比较低，在'''<font color="#ff8000">连接性Connectivity</font>'''方面也具有更加规则的结构。将'''<font color="#ff8000">神经形态芯片Neuromorphic chips</font>'''与大脑进行比较是一种非常粗糙的比较，类似于仅仅因为飞机有翅膀和尾巴就将它与鸟进行比较。事实上，当前最先进的人工智能在'''<font color="#ff8000">能耗效率Energy-efficiency</font>'''和计算效率方面距离人脑神经认知系统仍有较大差距，而神经形态工程只是一种通过从大脑机制中获得灵感来缩小这种差距的尝试，就像许多工程设计中都具有'''<font color="#ff8000">生物启发的特征Bio-inspired features</font>'''一样。<br />
<br />
===Democratic concerns===<br />
===公众担忧===<br />
Significant ethical limitations may be placed on neuromorphic engineering due to public perception.<ref name=":27">{{Cite report|url=https://ai100.stanford.edu/sites/g/files/sbiybj9861/f/ai_100_report_0831fnl.pdf|title=Artificial Intelligence and Life in 2030|author=2015 Study Panel|date=September 2016|work=One Hundred Year Study on Artificial Intelligence (AI100)|publisher=Stanford University}}</ref> Special [[Eurobarometer]] 382: Public Attitudes Towards Robots, a survey conducted by the European Commission, found that 60% of [[European Union]] citizens wanted a ban of robots in the care of children, the elderly, or the disabled. Furthermore, 34% were in favor of a ban on robots in education, 27% in healthcare, and 20% in leisure. The European Commission classifies these areas as notably “human.” The report cites increased public concern with robots that are able to mimic or replicate human functions. Neuromorphic engineering, by definition, is designed to replicate the function of the human brain.<ref name=":1">{{Cite web|url=http://ec.europa.eu/commfrontoffice/publicopinion/archives/ebs/ebs_382_en.pdf|title=Special Eurobarometer 382: Public Attitudes Towards Robots|last=European Commission|date=September 2012|website=European Commission}}</ref><br />
<br />
由于公众认知的相关忧虑，神经形态工程学可能会受到严重的伦理限制。<ref name=":27" />欧盟委员会进行的一项调查发现，60% 的欧盟公民希望禁止机器人参与照顾儿童、老人或残疾人的工作。此外，34% 的人支持禁止机器人用于教育，27% 的人支持禁止机器人用于医疗保健，20% 的人支持禁止机器人用于休闲。欧盟委员会将以上领域划入“人类”范畴。报告指出，公众越来越关注能够模仿或复制人类功能的机器人。神经形态工程，顾名思义，是为了复制人脑的功能而设计的。<ref name=":1" /><br />
<br />
The democratic concerns surrounding neuromorphic engineering are likely to become even more profound in the future. The European Commission found that EU citizens between the ages of 15 and 24 are more likely to think of robots as human-like (as opposed to instrument-like) than EU citizens over the age of 55. When presented an image of a robot that had been defined as human-like, 75% of EU citizens aged 15–24 said it corresponded with the idea they had of robots while only 57% of EU citizens over the age of 55 responded the same way. The human-like nature of neuromorphic systems, therefore, could place them in the categories of robots many EU citizens would like to see banned in the future.<ref name=":1" /><br />
<br />
围绕神经形态工程的公众担忧可能在未来变得更加深刻。欧盟委员会发现，相比于55岁以上的欧盟公民，15至24岁的欧盟公民更有可能认为机器人像人(而不是像仪器)。当看到一张“类人”机器人的图片时，年龄在15岁至24岁之间的欧盟公民中有75% 的人表示这符合他们对机器人的想法，而55岁以上的欧盟公民中只有57% 的人有同样的反应。因此，神经形态系统可能因为其类似人类的特性而被归入许多欧盟公民希望在未来禁止使用的机器人类别。<ref name=":1" /><br />
<br />
=== Personhood===<br />
=== 人格问题===<br />
As neuromorphic systems have become increasingly advanced, some scholars{{who|date=August 2021}} have advocated for granting [[personhood]] rights to these systems. If the brain is what grants humans their personhood, to what extent does a neuromorphic system have to mimic the human brain to be granted personhood rights? Critics of technology development in the [[Human Brain Project]], which aims to advance brain-inspired computing, have argued that advancement in neuromorphic computing could lead to machine consciousness or personhood.<ref name=":28">{{Cite journal|last=Aicardi|first=Christine|date=September 2018|title=Accompanying technology development in the Human Brain Project: From foresight to ethics management|journal=Futures|volume=102|pages=114–124|doi=10.1016/j.futures.2018.01.005|doi-access=free}}</ref> If these systems are to be treated as people, critics argue, then many tasks humans perform using neuromorphic systems, including the act of termination of neuromorphic systems, may be morally impermissible as these acts would violate the autonomy of the neuromorphic systems.<ref name=":29">{{Cite journal|last=Lim|first=Daniel|date=2014-06-01|title=Brain simulation and personhood: a concern with the Human Brain Project|journal=Ethics and Information Technology|language=en|volume=16|issue=2|pages=77–89|doi=10.1007/s10676-013-9330-5|s2cid=17415814|issn=1572-8439}}</ref><br />
<br />
随着神经形态系统的日益发展，一些学者主张赋予这些系统'''<font color="ff8000">人格Personhood</font>'''权。如果是大脑赋予了人类人格，那么在多大程度上模仿人类大脑的神经形态系统才能被赋予人格权利？“人类大脑计划”旨在推进以大脑为灵感的计算机技术发展，该计划的批评者认为，神经形态计算机技术的进步可能导致机器意识或人格的形成。<ref name=":28" />这些批评者认为，如果这些系统被当作人来对待，那么人类使用神经形态系统执行任务（包括终止神经形态系统）的行为，在道德上就可能是不被允许的，因为这些行为将违反神经形态系统的自主性。<ref name=":29" /><br />
<br />
<br />
==Dual use (military applications)==<br />
==军民两用技术（军事应用）==<br />
The [[Joint Artificial Intelligence Center]], a branch of the U.S. military, is a center dedicated to the procurement and implementation of AI software and neuromorphic hardware for combat use. Specific applications include smart headsets/goggles and robots. JAIC intends to rely heavily on neuromorphic technology to connect "every fighter every shooter" within a network of neuromorphic-enabled units.<br />
<br />
'''<font color="#ff8000">联合人工智能中心The Joint Artificial Intelligence Center</font>'''(JAIC),是美国军队的一个分支，专门从事采购和实施用于战斗的人工智能软件和神经形态硬件。具体应用包括智能耳机/护目镜和机器人。JAIC打算高度依赖神经形态技术，'''<font color="#32CD32">使用神经形态技术来连接神经形态单位网络中的“每个战士每个射手”</font>'''。<br />
<br />
<br />
==Legal considerations==<br />
==法律问题==<br />
Skeptics have argued that there is no way to apply the electronic personhood, the concept of personhood that would apply to neuromorphic technology, legally. In a letter signed by 285 experts in law, robotics, medicine, and ethics opposing a European Commission proposal to recognize “smart robots” as legal persons, the authors write, “A legal status for a robot can’t derive from the [[Natural person|Natural Person]] model, since the robot would then hold [[human rights]], such as the right to dignity, the right to its integrity, the right to remuneration or the right to citizenship, thus directly confronting the Human rights. This would be in contradiction with the [[Charter of Fundamental Rights of the European Union]] and the [[Convention for the Protection of Human Rights and Fundamental Freedoms]].”<ref name=":30">{{Cite web|url=http://www.robotics-openletter.eu/|title=Robotics Openletter {{!}} Open letter to the European Commission|language=fr-FR|access-date=2019-05-10}}</ref><br />
<br />
质疑者者认为，在法律上没有办法应用能够适用于神经形态技术的电子人格。在一封由285名法律、机器人技术、医学和伦理学专家的联名信中，作者们反对欧盟委员会提出的承认“智能机器人”为法人的提议。他们写道，“机器人的法律地位无法从'''<font color="#ff8000">自然人Natural Person</font>'''模型中推导出来，因为机器人将被赋予'''<font color="#ff8000">人权Human rights</font>'''，如尊严权、完整权、报酬权或公民权，从而直接面临人权问题。这将有悖于《欧联基本权利宪章》和《欧洲保障人权和根本自由公约》”。<ref name=":30" /><br />
<br />
===Ownership and property rights ===<br />
===所有权及财产权问题===<br />
There is significant legal debate around property rights and artificial intelligence. In ''Acohs Pty Ltd v. Ucorp Pty Ltd'', Justice Christopher Jessup of the [[Federal Court of Australia]] found that the [[source code]] for [[Material safety data sheets|Material Safety Data Sheets]] could not be [[Copyright law of Australia|copyrighted]] as it was generated by a [[software interface]] rather than a human author.<ref name=":31">{{Cite web|url=http://www.lavan.com.au/advice/intellectual_property_technology/copyright_in_source_code_and_digital_products|title=Copyright in source code and digital products|last=Lavan|website=Lavan|language=en|access-date=2019-05-10}}</ref> The same question may apply to neuromorphic systems: if a neuromorphic system successfully mimics a human brain and produces a piece of original work, who, if anyone, should be able to claim ownership of the work?<ref name=":32">{{cite journal |last1=Eshraghian|first1=Jason K. |title=Human Ownership of Artificial Creativity |journal=Nature Machine Intelligence |date=9 March 2020 |volume=2 |pages=157–160 |doi=10.1038/s42256-020-0161-x}}</ref><br />
<br />
法律界围绕财产权和人工智能有着重大争论。在Acohs Pty有限公司诉Ucorp Pty有限公司一案中，澳大利亚联邦法院的克里斯托弗·杰瑟普法官发现，版权保护不适用于材料安全数据表的'''<font color="#ff8000">源代码Source code</font>'''，因为它是由'''<font color="#ff8000">软件界面Software interface</font>'''生成而非人类工作者生成的。<ref name=":31" />同样的问题可能也适用于神经形态系统: 如果一个神经形态系统成功地模仿了人类的大脑，并产生了一部原创作品，那么该如何确认这部作品的所有权归属？<ref name=":32" /><br />
<br />
<br />
==Neuromemristive systems==<br />
==神经记忆电阻系统==<br />
Neuromemristive systems are a subclass of neuromorphic computing systems that focus on the use of [[memristors]] to implement [[neuroplasticity]]. While neuromorphic engineering focuses on mimicking biological behavior, neuromemristive systems focus on abstraction.<ref name=":33">{{Cite web|url=https://digitalops.sandia.gov/Mediasite/Play/a10cf6ceb55d47608bb8326dd00e46611d|title=002.08 N.I.C.E. Workshop 2014: Towards Intelligent Computing with Neuromemristive Circuits and Systems - Feb. 2014|website=digitalops.sandia.gov|access-date=2019-08-26}}</ref> For example, a neuromemristive system may replace the details of a [[Cerebral cortex|cortical]] microcircuit's behavior with an abstract neural network model.<ref name=":34">C. Merkel and D. Kudithipudi, "Neuromemristive extreme learning machines for pattern classification," ISVLSI, 2014.</ref><br />
<br />
神经记忆电阻系统是神经形态计算系统的一个亚类，主要研究利用'''<font color="ff8000">记忆电阻器Memristors</font>'''实现'''<font color="ff8000">神经可塑性Neuroplasticity</font>'''。神经形态工程的重点是模拟生物行为，而神经记忆电阻系统的重点是提取。<ref name=":33" />举个例子，一个神经记忆系统可能用抽象的神经网络模型替代'''<font color="ff8000">皮层Cortical</font>'''微电路的行为细节。<ref name=":34" /><br />
<br />
There exist several neuron inspired threshold logic functions<ref name="Maan 1–13" /> implemented with memristors that have applications in high level [[pattern recognition]] applications. Some of the applications reported recently include [[speech recognition]],<ref name=":35">{{Cite journal|title = Memristor pattern recogniser: isolated speech word recognition|journal = Electronics Letters|pages = 1370–1372|volume = 51|issue = 17|doi = 10.1049/el.2015.1428|first1 = A.K.|last1 = Maan|first2 = A.P.|last2 = James|first3 = S.|last3 = Dimitrijev|year = 2015|bibcode = 2015ElL....51.1370M|hdl = 10072/140989|s2cid = 61454815|url = https://semanticscholar.org/paper/48d3ab11ec6e213b62f11eedcfb7b7febb058674|hdl-access = free}}</ref> [[face recognition]]<ref name=":36">{{Cite journal|title = Memristive Threshold Logic Face Recognition|journal = Procedia Computer Science|date = 2014-01-01|pages = 98–103|volume = 41|series = 5th Annual International Conference on Biologically Inspired Cognitive Architectures, 2014 BICA|doi = 10.1016/j.procs.2014.11.090|first1 = Akshay Kumar|last1 = Maan|first2 = Dinesh S.|last2 = Kumar|first3 = Alex Pappachen|last3 = James|doi-access = free}}</ref> and [[object recognition]].<ref name=":37">{{Cite journal|title = Memristive Threshold Logic Circuit Design of Fast Moving Object Detection|journal = IEEE Transactions on Very Large Scale Integration (VLSI) Systems|date = 2015-10-01|issn = 1063-8210|pages = 2337–2341|volume = 23|issue = 10|doi = 10.1109/TVLSI.2014.2359801|first1 = A.K.|last1 = Maan|first2 = D.S.|last2 = Kumar|first3 = S.|last3 = Sugathan|first4 = A.P.|last4 = James|arxiv = 1410.1267|s2cid = 9647290}}</ref> They also find applications in replacing conventional digital logic gates.<ref name=":38">{{Cite journal|title = Resistive Threshold Logic|journal = IEEE Transactions on Very Large Scale Integration (VLSI) Systems|date = 2014-01-01|issn = 1063-8210|pages = 190–195|volume = 22|issue = 1|doi = 10.1109/TVLSI.2012.2232946|first1 = A.P.|last1 = James|first2 = L.R.V.J.|last2 = Francis|first3 = D.S.|last3 = Kumar|arxiv = 1308.0090|s2cid = 7357110}}</ref><ref name=":39">{{Cite journal|title = Threshold Logic Computing: Memristive-CMOS Circuits for Fast Fourier Transform and Vedic Multiplication|journal = IEEE Transactions on Very Large Scale Integration (VLSI) Systems|date = 2015-11-01|issn = 1063-8210|pages = 2690–2694|volume = 23|issue = 11|doi = 10.1109/TVLSI.2014.2371857|first1 = A.P.|last1 = James|first2 = D.S.|last2 = Kumar|first3 = A.|last3 = Ajayan|arxiv = 1411.5255|s2cid = 6076956}}</ref><br />
<br />
受神经元启发、使用记忆电阻器实现的阈值逻辑函数<ref name="Maan 1–13" />在高级'''<font color="#ff8000">模式识别Pattern recognition</font>'''中有着广泛的应用，最近报道中其应用包括'''<font color="#ff8000">语音识别Speech recognition<ref name=":35" /></font>'''、'''<font color="#ff8000">人脸识别Face recognition<ref name=":36" /></font>'''和'''<font color="#ff8000">物体识别Object recognition<ref name=":37" /></font>'''。阈值逻辑函数还可以用来取代传统的数字逻辑门。<ref name=":38" /><ref name=":39" /><br />
<br />
For ideal passive memristive circuits there is an exact equation (Caravelli-Traversa-[[Di Ventra]] equation) for the internal memory of the circuit:<ref name=":40">{{cite journal |last=Caravelli |display-authors=etal|arxiv=1608.08651 |title=The complex dynamics of memristive circuits: analytical results and universal slow relaxation |year=2017 |doi=10.1103/PhysRevE.95.022140 |pmid= 28297937 |volume=95 |issue= 2 |pages= 022140 |journal=Physical Review E|bibcode=2017PhRvE..95b2140C |s2cid=6758362}}</ref><br />
<br />
对于理想的无源记忆电路，电路的内部记忆可以用精确的方程(Caravelli-Traversa-Di Ventra方程) 来描述:<ref name=":40" /><br />
<br />
:<math> \frac{d}{dt} \vec{W} = \alpha \vec{W}-\frac{1}{\beta} (I+\xi \Omega W)^{-1} \Omega \vec S </math><br />
<br />
as a function of the properties of the physical memristive network and the external sources. In the equation above, <math>\alpha</math> is the "forgetting" time scale constant, <math> \xi=r-1</math> and <math>r=\frac{R_\text{off}}{R_\text{on}}</math> is the ratio of ''off'' and ''on'' values of the limit resistances of the memristors, <math> \vec S </math> is the vector of the sources of the circuit and <math>\Omega</math> is a projector on the fundamental loops of the circuit. The constant <math>\beta</math> has the dimension of a voltage and is associated to the properties of the [[memristor]]; its physical origin is the charge mobility in the conductor. The diagonal matrix and vector <math>W=\operatorname{diag}(\vec W)</math> and <math>\vec W</math> respectively, are instead the internal value of the memristors, with values between 0 and 1. This equation thus requires adding extra constraints on the memory values in order to be reliable.<br />
<br />
Caravelli-Traversa-Di Ventra方程是描述物理记忆网络和外部源性质的函数。在上述方程中，<math>\alpha</math>是“遗忘”时间尺度常数，<math>\xi=r-1</math>，<math>r =\frac{R\text_{off}}{R_\text{on}}</math>是记忆电阻器off状态和on状态极限电阻值之比，<math>\vec S</math>是电路源的矢量，<math>\Omega</math>是电路基本环路的投影。常数<math>\beta</math>具有电压的量纲，与记忆电阻器的特性有关；其物理原型是导体中的电荷迁移率。对角矩阵和向量 <math>W=\operatorname{diag}(\vec W)</math>和<math>\vec W</math> '''<font color="#32CD32">分别是记忆电阻器的内阻</font>'''，值在0到1之间。因此，这个等式需要在'''<font color="32CD32">内存值</font>'''上添加额外约束以保证可靠性。<br />
<br />
<br />
<br />
==See also==<br />
==相关词条==<br />
{{Columns-list|colwidth=18em|<br />
* [[AI accelerator (computer hardware)|AI accelerator]]<br />
* [[Artificial brain]]<br />
* [[Biomorphic]]<br />
* [[Cognitive computer]]<br />
* [[Computation and Neural Systems]]<br />
* [[Differentiable programming]]<br />
* [[Event camera]]<br />
* [[Neurorobotics]]<br />
* [[Optical flow sensor]]<br />
* [[Physical neural network]]<br />
* [[SpiNNaker]]<br />
* [[SyNAPSE]]<br />
* [[Retinomorphic sensor]]<br />
* [[Vision chip]]<br />
* [[Vision processing unit]]<br />
* [[Zeroth (software)]]<br />
* [[Hardware for artificial intelligence]]<br />
}}<br />
<br />
{{Portal bar|Electronics}}<br />
<br />
==References==<br />
{{Reflist|40em}}<br />
<br />
==External links==<br />
<!--======================== {{No more links}} ============================<br />
| PLEASE BE CAUTIOUS IN ADDING MORE LINKS TO THIS ARTICLE. Wikipedia |<br />
| is not a collection of links nor should it be used for advertising. |<br />
| |<br />
| Excessive or inappropriate links WILL BE DELETED. |<br />
| See [[Wikipedia:External links]] & [[Wikipedia:Spam]] for details. |<br />
| |<br />
| If there are already plentiful links, please propose additions or |<br />
| replacements on this article's discussion page, or submit your link |<br />
| to the relevant category at the Open Directory Project (dmoz.org) |<br />
| and link back to that category using the {{dmoz}} template. |<br />
======================={{No more links}}=============================--><br />
{{Commons category|Neuromorphic Engineering}}<br />
*[https://web.archive.org/web/20150727034331/http://ine-web.org/workshops/workshops-overview Telluride Neuromorphic Engineering Workshop]<br />
*[https://archive.today/20130115190057/http://capocaccia.ethz.ch/ CapoCaccia Cognitive Neuromorphic Engineering Workshop]<br />
*[https://web.archive.org/web/20190716132350/http://www.ine-web.org/ Institute of Neuromorphic Engineering]<br />
*[http://www.ine-news.org/ INE news site].<br />
*[http://www.frontiersin.org/neuromorphic_engineering Frontiers in Neuromorphic Engineering Journal]<br />
*[http://www.cns.caltech.edu/ Computation and Neural Systems] department at the [[California Institute of Technology]].<br />
*[http://www.humanbrainproject.eu/ Human Brain Project official site]<br />
*[https://www.the-scientist.com/features/building-a-silicon-brain-65738 Building a Silicon Brain:] Computer chips based on biological neurons may help simulate larger and more-complex brain models. May 1, 2019. SANDEEP RAVINDRAN<br />
<br />
<br />
*Telluride Neuromorphic Engineering Workshop<br />
*CapoCaccia Cognitive Neuromorphic Engineering Workshop<br />
*Institute of Neuromorphic Engineering<br />
*INE news site.<br />
*Frontiers in Neuromorphic Engineering Journal<br />
*Computation and Neural Systems department at the California Institute of Technology.<br />
*Human Brain Project official site<br />
*Building a Silicon Brain: Computer chips based on biological neurons may help simulate larger and more-complex brain models. May 1, 2019. SANDEEP RAVINDRAN<br />
<br />
=<nowiki>外部链接</nowiki>= <br />
*碲化物神经形态工程工作室<br />
*CapoCaccia认知神经形态工程工作室<br />
*神经形态工程研究所<br />
*INE新闻站点。<br />
*《神经形态工程学前沿》<br />
*加州理工学院计算与神经系统系。<br />
*人脑项目官方网站<br />
*打造硅脑: 基于生物神经元的计算机芯片可能有助于模拟更大、更复杂的大脑模型。2019年5月1日。SANDEEP RAVINDRAN.<br />
<br />
{{Differentiable computing}}<br />
{{Authority control}}<br />
<br />
[[Category:Electrical engineering]]<br />
[[Category:Neuroscience]]<br />
[[分类:AI accelerators|*]]<br />
[[Category:Artificial intelligence]]<br />
[[Category:Robotics]]<br />
<br />
Category:Electrical engineering<br />
Category:Neuroscience<br />
*<br />
Category:Artificial intelligence<br />
Category:Robotics<br />
<br />
类别: 电气工程类别: 神经科学 * 类别: 人工智能类别: 机器人<br />
<br />
<noinclude><br />
<br />
<small>This page was moved from [[wikipedia:en:Neuromorphic engineering]]. Its edit history can be viewed at [[神经形态计算/edithistory]]</small></noinclude><br />
<br />
[[Category:待整理页面]]<br />
<br />
<references /></div>Shenky20https://wiki.swarma.org/index.php?title=%E7%A5%9E%E7%BB%8F%E5%BD%A2%E6%80%81%E8%AE%A1%E7%AE%97&diff=30753神经形态计算2022-04-28T11:58:19Z<p>Shenky20：/* 神经记忆电阻系统 */</p>
<hr />
<div><br />
此词条由神经动力学模型读书会词条梳理志愿者Shenky20翻译审校，未经专家审校，带来阅读不便，请见谅。<br />
<br />
{{Use American English|date = January 2019}}<br />
{{Use mdy dates|date = January 2019}}<br />
<br />
'''Neuromorphic engineering''', also known as '''neuromorphic computing''',<ref name=":3">{{Cite journal | doi = 10.1145/2601069| title = Neuromorphic computing gets ready for the (really) big time| journal = [[Communications of the ACM]]| volume = 57| issue = 6| year = 2014| last1 = Monroe | first1 = D. | pages = 13–15| s2cid = 20051102}}</ref><ref name=":4">{{Cite journal | doi = 10.1088/0957-4484/21/17/175202| pmid = 20368686| title = Nanotube devices based crossbar architecture: Toward neuromorphic computing| journal = Nanotechnology| volume = 21| issue = 17| pages = 175202| year = 2010| last1 = Zhao | first1 = W. S. | last2 = Agnus | first2 = G. | last3 = Derycke | first3 = V. | last4 = Filoramo | first4 = A. | last5 = Bourgoin | first5 = J. -P. | last6 = Gamrat | first6 = C. | bibcode = 2010Nanot..21q5202Z| url = https://zenodo.org/record/3428659}}</ref><ref name="humanbrainproject">{{YouTube|id=6RoiZ90mGfw|title=The Human Brain Project SP 9: Neuromorphic Computing Platform}}</ref> is the use of [[very-large-scale integration]] (VLSI) systems containing electronic [[analog circuit]]s to mimic neuro-biological architectures present in the nervous system. A neuromorphic computer/chip is any device that uses physical artificial neurons (made from silicon) to do computations.<ref name=":5">{{cite journal|last1=Mead|first1=Carver|title=Neuromorphic electronic systems|journal=Proceedings of the IEEE|date=1990|volume=78|issue=10|pages=1629–1636|doi=10.1109/5.58356|url=https://authors.library.caltech.edu/53090/1/00058356.pdf}}</ref><ref name=":2" /> In recent times, the term ''neuromorphic'' has been used to describe analog, digital, mixed-mode analog/digital VLSI, and software systems that implement models of [[neural system]]s (for [[perception]], [[motor control]], or [[multisensory integration]]). The implementation of neuromorphic computing on the hardware level can be realized by oxide-based [[memristor]]s,<ref name="Maan 1–13">{{Cite journal|last1=Maan|first1=A. K.|last2=Jayadevi|first2=D. A.|last3=James|first3=A. P.|date=2016-01-01|title=A Survey of Memristive Threshold Logic Circuits|journal=IEEE Transactions on Neural Networks and Learning Systems|volume=PP|issue=99|pages=1734–1746|doi=10.1109/TNNLS.2016.2547842|pmid=27164608|issn=2162-237X|arxiv=1604.07121|bibcode=2016arXiv160407121M|s2cid=1798273}}</ref> spintronic memories, threshold switches, and [[transistor]]s.<ref name=":6">{{Cite journal|title = Mott Memory and Neuromorphic Devices|journal = Proceedings of the IEEE|date = 2015-08-01|issn = 0018-9219|pages = 1289–1310|volume = 103|issue = 8|doi = 10.1109/JPROC.2015.2431914|first1 = You|last1 = Zhou|first2 = S.|last2 = Ramanathan|s2cid = 11347598|url = https://zenodo.org/record/895565}}</ref><ref name=":2">{{Cite document|title=Neuromorphic Circuits With Neural Modulation Enhancing the Information Content of Neural Signaling {{!}} International Conference on Neuromorphic Systems 2020|language=EN|doi=10.1145/3407197.3407204|s2cid=220794387}}</ref> Training software-based neuromorphic systems of spiking neural networks can be achieved using error backpropagation, e.g., using [[Python (programming language)|Python]] based frameworks such as snnTorch,<ref name=":7">{{cite journal |last1=Eshraghian|first1=Jason K.|last2=Ward|first2=Max|last3=Neftci |first3=Emre|last4=Wang|first4=Xinxin|last5=Lenz|first5=Gregor|last6=Dwivedi|first6=Girish|last7=Bennamoun|first7=Mohammed|last8=Jeong|first8=Doo Seok|last9=Lu|first9=Wei D.|title=Training Spiking Neural Networks Using Lessons from Deep Learning |date=1 October 2021 |arxiv=2109.12894 }}</ref> or using canonical learning rules from the biological learning literature, e.g., using BindsNet.<ref name=":8">{{Cite web | url=https://github.com/Hananel-Hazan/bindsnet | title=Hananel-Hazan/bindsnet: Simulation of spiking neural networks (SNNs) using PyTorch.| date=31 March 2020}}</ref><br />
<br />
<br />
'''<font color="#ff8000">神经形态工程Neuromorphic engineering</font>'''(也称为'''<font color="#ff8000">神经形态计算Neuromorphic computing</font>''')<ref name=":3" /><ref name=":4" /><ref name="humanbrainproject" />是指使用包含电子'''<font color="#ff8000">模拟电路Analog circuit</font>'''的'''<font color="#ff8000">超大规模集成电路Very-large-scale integration</font>'''系统来模拟神经系统中生理结构的研究方法。神经形态计算机或神经形态芯片包括任何使用由硅制成的人造神经元进行计算的设备。<ref name=":5" /><ref name=":2" />近年来，神经形态学（neuromorphic）这个术语被用来描述能够实现'''<font color="#ff8000">神经系统Neural system</font>'''模型功能(如'''<font color="#ff8000">感知Perception</font>'''、'''<font color="#ff8000">运动控制Motor control</font>'''，'''<font color="#ff8000">多感官整合Multisensory integration</font>'''等)的模拟、数字、模拟/数字混合模式超大规模集成电路和软件系统。神经形态计算的硬件实现可以通过基于氧化物的'''<font color="#ff8000">记忆电阻器Memristor</font>'''、<ref name="Maan 1–13" />自旋电子存储器、阈值开关和'''<font color="#ff8000">晶体管Transistor</font>'''来实现。<ref name=":6" /><ref name=":2" />对基于软件的脉冲神经网络系统的训练可以通过误差反向传播机制来实现，例如，使用snnTorch等基于Python的框架，<ref name=":7" />或使用BindsNet等典型的受生物启发的学习模式。<ref name=":8" /><br />
<br />
A key aspect of neuromorphic engineering is understanding how the morphology of individual neurons, circuits, applications, and overall architectures creates desirable computations, affects how information is represented, influences robustness to damage, incorporates learning and development, adapts to local change (plasticity), and facilitates evolutionary change.<br />
<br />
<br />
神经形态工程领域的一个关键问题，就是理解'''<font color="#32CD32">单个神经元形态、神经回路、应用和整体结构</font>'''如何产生理想的计算，如何影响信息的表示和对破坏的鲁棒性，如何整合学习和发展，如何适应局部变化(可塑性) 并促进逐渐发展的变化。<br />
<br />
Neuromorphic engineering is an interdisciplinary subject that takes inspiration from [[biology]], [[physics]], [[mathematics]], [[computer science]], and [[electronic engineering]]<ref name=":2" /> to design artificial neural systems, such as [[Machine vision|vision systems]], head-eye systems, auditory processors, and autonomous robots, whose physical architecture and design principles are based on those of biological nervous systems.<ref name=":9">{{Cite journal | doi = 10.1155/2012/705483| title = Qualitative Functional Decomposition Analysis of Evolved Neuromorphic Flight Controllers| journal = Applied Computational Intelligence and Soft Computing| volume = 2012| pages = 1–21| year = 2012| last1 = Boddhu | first1 = S. K. | last2 = Gallagher | first2 = J. C. | doi-access = free}}</ref> It was developed by [[Carver Mead]]<ref name=":10">{{cite web|title=carver mead website|last1=Mead|first1=Carver|url=http://www.carvermead.caltech.edu/index.html|website=carvermead}}</ref> in the late 1980s.<br />
<br />
<br />
神经形态工程是以'''<font color="#ff8000">生物学Biology</font>'''、'''<font color="#ff8000">物理学Physics</font>'''、'''<font color="#ff8000">数学Mathematics</font>'''、'''<font color="#ff8000">计算机科学Computer science</font>'''和'''<font color="#ff8000">电子工程Electronic engineering<ref name=":2" /></font>'''等学科为基础，设计人工神经系统（如'''<font color="#ff8000">视觉系统Vision systems</font>'''、头眼系统、听觉处理器以及物理结构和设计原则都受启发于生物神经系统的自主机器人）的一门交叉学科。<ref name=":9" />20世纪80年代后期，卡弗·米德极大地推动了神经形态工程领域的发展。<ref name=":10" /><br />
<br />
==Neurological inspiration==<br />
==神经生物学启发==<br />
Neuromorphic engineering is set apart by the inspiration it takes from what we know about the structure and operations of the [[brain]]. Neuromorphic engineering translates what we know about the brain's function into computer systems. Work has mostly focused on replicating the analog nature of [[biological computation]] and the role of [[neuron]]s in [[cognition]].<br />
<br />
<br />
神经形态工程形成的灵感来源于现已知的'''<font color="#ff8000">大脑Brain</font>'''结构及运作机制知识，它将我们对大脑功能的了解用于研究和优化计算机系统。该领域工作主要集中于对'''<font color="#ff8000">生物计算Biological computation</font>'''模拟特性和'''<font color="#ff8000">神经元Neuron</font>'''在'''<font color="#ff8000">认知Cognition</font>'''中所发挥作用的复刻。<br />
<br />
The biological processes of neurons and their [[synapse]]s are dauntingly complex, and thus very difficult to artificially simulate. A key feature of biological brains is that all of the processing in neurons use analog chemical signals. This makes it hard to replicate brains in computers because the current generation of computers is completely digital. However, the characteristics of these parts can be abstracted into mathematical functions that closely capture the essence of the neuron's operations.<br />
<br />
神经元及其'''<font color="#ff8000">突触Synapse</font>'''的生理学过程极其复杂，难以进行人工模拟。神经元在所有处理过程中都使用化学模拟信号，这是大脑的一个关键生理学特征。这个特征大大增加了在计算机中复制大脑的难度，因为目前的计算机是完全数字化的。然而，部分特征可以抽象为数学函数，这些函数能够紧密捕捉神经元运行的本质。<br />
<br />
The goal of neuromorphic computing is not to perfectly mimic the brain and all of its functions, but instead to extract what is known of its structure and operations to be used in a practical computing system. No neuromorphic system will claim nor attempt to reproduce every element of neurons and synapses, but all adhere to the idea that computation is highly [[distributed processing|distributed]] throughout a series of small computing elements analogous to a neuron. While this sentiment is standard, researchers chase this goal with different methods.<ref name=":11">{{Cite journal | doi = 10.1088/1741-2560/13/5/051001| title = Large-scale neuromorphic computing systems| journal = Journal of Neural Engineering| volume = 13| pages = 1–15| year = 2016| last1 = Furber | first1 = Steve| issue = 5| pmid = 27529195| bibcode = 2016JNEng..13e1001F| doi-access = free}}</ref><br />
<br />
神经形态计算的目标不是完美地模拟大脑及其所有功能，而是利用已知的大脑结构和运转机制来研发或优化实际的计算系统。任何神经形态学系统都不会声称或试图复制神经元和突触中的每一个元素，但所有人都一致认可将计算高度'''<font color="#ff8000">分散Distribute</font>'''于一系列类似于神经元的小型计算元素的理念。研究人员用不同的方法来追求这一普遍目标。<ref name=":11" /><br />
<br />
<br />
==Examples==<br />
==代表性成果==<br />
As early as 2006, researchers at [[Georgia Tech]] published a field programmable neural array.<ref name=":12">{{Cite book|title = A field programmable neural array|last1 = Farquhar|first1 = Ethan|date = May 2006|journal = IEEE International Symposium on Circuits and Systems|pages = 4114–4117|last2 = Hasler|first2 = Paul.|doi = 10.1109/ISCAS.2006.1693534|isbn = 978-0-7803-9389-9|s2cid = 206966013}}</ref> This chip was the first in a line of increasingly complex arrays of floating gate transistors that allowed programmability of charge on the gates of [[MOSFET]]s to model the channel-ion characteristics of neurons in the brain and was one of the first cases of a silicon programmable array of neurons.<br />
<br />
早在2006年，佐治亚理工学院的研究人员就研发出了一种现场可编程神经阵列。<ref name=":12" />在此之后，一系列越来越复杂的浮栅晶体管阵列被成功研发出来，这些晶体管阵列可以通过在'''<font color="#ff8000">金属-氧化物半导体效应晶体管MOSFET</font>'''的栅极上编程来模拟大脑中神经元的离子通道特性，这也是以硅为主要材料的可编程神经元阵列的最早成功案例之一。<br />
<br />
In November 2011, a group of [[MIT]] researchers created a computer chip that mimics the analog, ion-based communication in a synapse between two neurons using 400 transistors and standard [[CMOS]] manufacturing techniques.<ref name=":13">{{cite web|title=MIT creates "brain chip"|url=http://www.extremetech.com/extreme/105067-mit-creates-brain-chip|access-date=4 December 2012}}</ref><ref name="Neuromorphic silicon paper">{{cite journal|title=Neuromorphic silicon neurons and large-scale neural networks: challenges and opportunities|doi=10.3389/fnins.2011.00108|pmid=21991244|pmc=3181466|volume=5|pages=108|journal=Frontiers in Neuroscience|year=2011|last1=Poon|first1=Chi-Sang|last2=Zhou|first2=Kuan|doi-access=free}}</ref><br />
<br />
2011年11月，麻省理工学院的一组研究人员研发出一种计算机芯片，该芯片上使用标准的'''<font color="#ff8000">互补金属氧化物半导体CMOS</font>'''制造技术集成了400个晶体管，能够模拟神经元间突触中基于离子的通讯。<ref name=":13" /><ref name="Neuromorphic silicon paper" /><br />
<br />
In June 2012, [[spintronic]] researchers at [[Purdue University]] presented a paper on the design of a neuromorphic chip using [[Spin valve|lateral spin valve]]s and [[memristor]]s. They argue that the architecture works similarly to neurons and can therefore be used to test methods of reproducing the brain's processing. In addition, these chips are significantly more energy-efficient than conventional ones.<ref name="Spin Devices Prop">{{Cite arXiv|title=Proposal For Neuromorphic Hardware Using Spin Devices|eprint=1206.3227|last1=Sharad|first1=Mrigank|last2=Augustine|first2=Charles|last3=Panagopoulos|first3=Georgios|last4=Roy|first4=Kaushik|class=cond-mat.dis-nn|year=2012}}</ref><br />
<br />
2012年6月，普渡大学的'''<font color="#ff8000">自旋电子学Spintronic</font>'''研究人员发表了一篇关于利用'''<font color="#ff8000">侧向自旋阀Lateral spin valves</font>'''和'''<font color="#ff8000">记忆电阻器Memristors</font>'''设计神经形态芯片的论文。他们认为，这种芯片结构的工作原理与神经元相似，因此可以用于大脑运行机制的复刻方法的测试。此外，这些芯片在能耗方面明显优于传统芯片。<ref name="Spin Devices Prop" /><br />
<br />
Research at [[HP Labs]] on Mott memristors has shown that while they can be non-[[Volatile memory|volatile]], the volatile behavior exhibited at temperatures significantly below the [[phase transition]] temperature can be exploited to fabricate a [[neuristor]],<ref name=":0" /> a biologically-inspired device that mimics behavior found in neurons.<ref name=":0">{{Cite journal | doi = 10.1038/nmat3510| pmid = 23241533| title = A scalable neuristor built with Mott memristors| journal = Nature Materials| volume = 12| issue = 2| pages = 114–7| year = 2012| last1 = Pickett | first1 = M. D. | last2 = Medeiros-Ribeiro | first2 = G. | last3 = Williams | first3 = R. S. | bibcode = 2013NatMa..12..114P| s2cid = 16271627| url = https://semanticscholar.org/paper/b6ba6f496ace2b947f111059663e76bb60e9efeb}}</ref> In September 2013, they presented models and simulations that show how the spiking behavior of these neuristors can be used to form the components required for a [[Turing machine]].<ref name=":14">{{cite journal|doi=10.1088/0957-4484/24/38/384002|title=Phase transitions enable computational universality in neuristor-based cellular automata|author1=Matthew D Pickett|author2=R Stanley Williams|name-list-style=amp|date=September 2013|publisher=IOP Publishing Ltd|journal=Nanotechnology|volume=24|issue=38|pmid=23999059|bibcode=2013Nanot..24L4002P|at=384002}}</ref><br />
<br />
'''<font color="#ff8000">惠普实验室HP labs</font>'''在莫特记忆电阻器上的研究表明，尽管它们可以是非'''<font color="#ff8000">易失性Volatile</font>'''的，但是在'''<font color="#ff8000">相变Phase transition</font>'''温度以下时表现出的易失性行为可以被用来制造'''<font color="#ff8000">类神经元电阻器Neuristor</font>'''(一种生物学启发的模仿神经元行为的硬件)<ref name=":0" />。2013年9月，他们通过模型和仿真展示了这些类神经元电阻器的脉冲行为如何产生'''<font color="#ff8000">图灵机Turing machine</font>'''的所需元素。<ref name=":14" /><br />
<br />
[[Neurogrid]], built by ''Brains in Silicon'' at [[Stanford University]],<ref name=":15">{{cite journal|last1=Boahen|first1=Kwabena|title=Neurogrid: A Mixed-Analog-Digital Multichip System for Large-Scale Neural Simulations|journal=Proceedings of the IEEE|date=24 April 2014|volume=102|issue=5|pages=699–716|doi=10.1109/JPROC.2014.2313565|s2cid=17176371}}</ref> is an example of hardware designed using neuromorphic engineering principles. The circuit board is composed of 16 custom-designed chips, referred to as NeuroCores. Each NeuroCore's analog circuitry is designed to emulate neural elements for 65536 neurons, maximizing energy efficiency. The emulated neurons are connected using digital circuitry designed to maximize spiking throughput.<ref name=":16">{{cite journal|doi=10.1038/503022a|pmid = 24201264|title = Neuroelectronics: Smart connections|journal = Nature|volume = 503|issue = 7474|pages = 22–4|year = 2013|last1 = Waldrop|first1 = M. Mitchell|bibcode = 2013Natur.503...22W|doi-access = free}}</ref><ref name=":17">{{cite journal|doi=10.1109/JPROC.2014.2313565|title = Neurogrid: A Mixed-Analog-Digital Multichip System for Large-Scale Neural Simulations|journal = Proceedings of the IEEE|volume = 102|issue = 5|pages = 699–716|year = 2014|last1 = Benjamin|first1 = Ben Varkey|last2 = Peiran Gao|last3 = McQuinn|first3 = Emmett|last4 = Choudhary|first4 = Swadesh|last5 = Chandrasekaran|first5 = Anand R.|last6 = Bussat|first6 = Jean-Marie|last7 = Alvarez-Icaza|first7 = Rodrigo|last8 = Arthur|first8 = John V.|last9 = Merolla|first9 = Paul A.|last10 = Boahen|first10 = Kwabena|s2cid = 17176371}}</ref><br />
<br />
'''<font color="#ff8000">神经栅格Neurogrid<ref name=":15" /></font>'''是由斯坦福大学Brains in Silicon公司研发的、使用神经形态工程原理设计的硬件。该电路板由16个定制设计的芯片组成(NeuroCores)。在设计中，每个NeuroCore芯片的模拟电路对65536个神经元的神经元素进行模拟，以最大限度地提高能量效率。模拟出的神经元通过设计的数字电路连接，以最大化脉冲吞吐量。<ref name=":16" /><ref name=":17" /><br />
<br />
A research project with implications for neuromorphic engineering is the [[Human Brain Project]] that is attempting to simulate a complete human brain in a supercomputer using biological data. It is made up of a group of researchers in neuroscience, medicine, and computing.<ref name=":18">{{cite web|title=Involved Organizations|url=http://www.humanbrainproject.eu/partners.html|access-date=22 February 2013|url-status=dead|archive-url=https://web.archive.org/web/20130302142627/http://www.humanbrainproject.eu/partners.html|archive-date=2 March 2013}}</ref> [[Henry Markram]], the project's co-director, has stated that the project proposes to establish a foundation to explore and understand the brain and its diseases, and to use that knowledge to build new computing technologies. The three primary goals of the project are to better understand how the pieces of the brain fit and work together, to understand how to objectively diagnose and treat brain diseases, and to use the understanding of the human brain to develop neuromorphic computers. That the simulation of a complete human brain will require a supercomputer a thousand times more powerful than today's encourages the current focus on neuromorphic computers.<ref name=":19">{{cite web|title=Human Brain Project|url=http://www.humanbrainproject.eu|access-date=22 February 2013}}</ref> $1.3 billion has been allocated to the project by The [[European Commission]].<ref name=":20">{{cite web|title=The Human Brain Project and Recruiting More Cyberwarriors|url=http://www.marketplace.org/topics/tech/human-brain-project-and-recruiting-more-cyberwarriors|access-date=22 February 2013|date=January 29, 2013}}</ref><br />
<br />
'''<font color="#ff8000">人类大脑计划Human Brain Project</font>'''对神经形态工程具有较大影响，其主要任务是尝试用生物数据在超级计算机中模拟完整的人脑。人类大脑计划由神经科学、医学和计算机科学背景的研究人员组成。<ref name=":18" />该项目的联合主管亨利•马克拉姆(Henry Markram)表示，人类大脑计划的目的是建立一个探索和了解脑科学和脑疾病知识的基础，并利用这些知识来构建更先进的计算机技术。这个项目的三个主要目标分别是: 更好地理解大脑的各个部分是如何相互配合协同工作的; 理解如何客观地诊断和治疗脑部疾病; 以及利用对人类大脑的理解来开发神经形态计算机。模拟一个完整的人类大脑需要一台比现在强大一千倍的超级计算机，这不断激发着对神经形态计算机领域的研究兴趣。<ref name=":19" />欧盟委员会已经向人类大脑计划拨款13亿美元。<ref name=":20" /><br />
<br />
Other research with implications for neuromorphic engineering involves the [[BRAIN Initiative]]<ref name="economist">[https://www.economist.com/news/science-and-technology/21582495-computers-will-help-people-understand-brains-better-and-understanding-brains Neuromorphic computing: The machine of a new soul], The Economist, 2013-08-03</ref> and the [[TrueNorth]] chip from [[IBM]].<ref name=":21">{{cite journal|last1=Modha|first1=Dharmendra|title=A million spiking-neuron integrated circuit with a scalable communication network and interface|journal=Science|date=Aug 2014|volume=345|issue=6197|pages=668–673|doi=10.1126/science.1254642|pmid=25104385|bibcode=2014Sci...345..668M|s2cid=12706847}}</ref> Neuromorphic devices have also been demonstrated using nanocrystals, nanowires, and conducting polymers.<ref name=":22">{{Cite web|url=http://jessamynfairfield.com/wp-content/uploads/2017/03/PWMar17Fairfield.pdf|title=Smarter Machines|last=Fairfield|first=Jessamyn|date=March 1, 2017}}</ref><br />
<br />
其他与神经形态工程有关的研究还包括'''<font color="#ff8000">脑计划BRAIN initiative</font>'''，<ref name="economist" />和IBM研发的'''<font color="#ff8000">TrueNorth</font>'''芯片。<ref name=":21" />使用纳米晶体、纳米线和导电聚合物也能够用于制造神经形态学硬件。<ref name=":22" /><br />
<br />
[[Intel]] unveiled its neuromorphic research chip, called “[[Intel Loihi|Loihi]]”, in October 2017. The chip uses an asynchronous [[spiking neural network]] (SNN) to implement adaptive self-modifying event-driven fine-grained parallel computations used to implement learning and inference with high efficiency.<ref name=":23">{{cite journal |last1=Davies |first1=Mike |title=Loihi: A Neuromorphic Manycore Processor with On-Chip Learning |journal=IEEE Micro |date=January 16, 2018 |volume=38 |issue=1 |pages=82–99 |display-authors=etal|doi=10.1109/MM.2018.112130359 |s2cid=3608458 }}</ref><ref name="Morris2017">{{cite web |last1=Morris |first1=John |title=Why Intel built a neuromorphic chip |url=https://www.zdnet.com/article/why-intel-built-a-neuromorphic-chip/ |website=ZDNet |access-date=17 August 2018 |language=en}}</ref><br />
<br />
2017年10月，英特尔发布了神经形态芯片'''<font color="#ff8000">Loihi</font>'''。该芯片采用异步'''<font color="#ff8000">脉冲神经网络Spiking neural network</font>'''实现了自适应、自修改、事件驱动的细粒度并行计算，实现了高效的学习和推理。<ref name=":23" /><ref name="Morris2017" /><br />
<br />
[[IMEC]], a Belgium-based nanoelectronics research center, demonstrated the world's first self-learning neuromorphic chip. The brain-inspired chip, based on OxRAM technology, has the capability of self-learning and has been demonstrated to have the ability to compose music.<ref name=":24">{{cite web |title=Imec demonstrates self-learning neuromorphic chip that composes music |url=https://www.imec-int.com/en/articles/imec-demonstrates-self-learning-neuromorphic-chip-that-composes-music |website=IMEC International |access-date=1 October 2019}}</ref> IMEC released the 30-second tune composed by the prototype. The chip was sequentially loaded with songs in the same time signature and style. The songs were old Belgian and French flute minuets, from which the chip learned the rules at play and then applied them.<ref name=":25">{{cite web|last1=Bourzac|first1=Katherine|title=A Neuromorphic Chip That Makes Music|url=https://spectrum.ieee.org/a-neuromorphic-chip-that-makes-music|url-status=live|access-date=1 October 2019|website=IEEE Spectrum|date=May 23, 2017}}</ref><br />
<br />
比利时的'''<font color="#ff8000">微电子研究中心IMEC</font>'''研发了世界上首个自学习神经形态芯片。这种基于'''<font color="#32CD32"> OxRAM(filamentary-oxide-based resistive memory technology)</font>''' 技术的大脑启发芯片具有自学习能力，并且已被证明具有创作音乐的能力。<ref name=":24" />IMEC发布了由芯片原型机谱写的30秒旋律。向芯片加载一系列特征、风格统一的歌曲（古代比利时和法国长笛小步舞曲），芯片从中学习相关规则并将其应用于创作。<ref name=":25" /><br />
<br />
[[Blue Brain Project|The Blue Brain Project]], led by Henry Markram, aims to build biologically detailed digital reconstructions and simulations of the mouse brain. The Blue Brain Project has created in silico models of rodent brains, while attempting to replicate as many details about its biology as possible. The supercomputer-based simulations offer new perspectives on understanding the structure and functions of the brain.<br />
<br />
由亨利·马克拉姆领导的'''<font color="#ff8000">蓝脑计划The Blue Brain Project</font>'''旨在建立小鼠大脑生理学细节的数字重建和模拟。蓝脑计划已经建立了啮齿动物大脑的电子模型，同时进行着尽可能多地复制其生理学细节的尝试。基于超级计算机的模拟为理解大脑的结构和功能提供了新的视角。<br />
<br />
The European Union funded a series of projects at the University of Heidelberg, which led to the development of <br />
[[BrainScaleS]] (brain-inspired multiscale computation in neuromorphic hybrid systems), a hybrid analog [[neuromorphic]] supercomputer located at Heidelberg University, Germany. It was developed as part of the [[Human Brain Project]] neuromorphic computing platform and is the complement to the [[SpiNNaker]] supercomputer (which is based on digital technology). The architecture used in BrainScaleS mimics biological neurons and their connections on a physical level; additionally, since the components are made of silicon, these model neurons operate on average 864 times (24 hours of real time is 100 seconds in the machine simulation) that of their biological counterparts.<ref name=":26">{{Cite web|date=2016-03-21|title=Beyond von Neumann, Neuromorphic Computing Steadily Advances|url=https://www.hpcwire.com/2016/03/21/lacking-breakthrough-neuromorphic-computing-steadily-advance/|access-date=2021-10-08|website=HPCwire|language=en-US}}</ref><br />
<br />
欧盟资助了海德堡大学一系列促进BrainScaleS(神经形态混合系统中受大脑启发的多尺度计算)发展的项目，这是一台位于德国海德堡大学的混合模拟'''<font color="#ff8000">神经形态Neuromorphic</font>'''超级计算机。它是作为人类大脑计划中神经形态计算平台的一部分而开发的，是'''<font color="#ff8000">SpiNNaker</font>'''超级计算机(基于数字技术)的补充。BrainScaleS中使用的体系架构模拟了生物神经元及其在物理层面上的连接；此外，由于这些组件是由硅制成的，这些模型神经元平均运行速度是生物神经元的864倍，这意味着在机器模拟中，24小时的实时时间仅为100秒。<ref name=":26" /><br />
<br />
===Neuromorphic sensors===<br />
===神经形态传感器===<br />
The concept of neuromorphic systems can be extended to sensors (not just to computation). An example of this applied to detecting [[light]] is the [[retinomorphic sensor]] or, when employed in an array, the [[event camera]].<br />
<br />
神经形态系统的概念可以扩展到传感器(而不仅仅是计算单元)。一个用于检测光线的例子是'''<font color="#ff8000">类视网膜传感器Retinomorphic sensor</font>'''，或者'''<font color="#ff8000">事件摄像机Event camera</font>阵列。'''<br />
<br />
<br />
==Ethical considerations ==<br />
==伦理问题==<br />
While the interdisciplinary concept of neuromorphic engineering is relatively new, many of the same ethical considerations apply to neuromorphic systems as apply to [[human-like machines]] and [[artificial intelligence]] in general. However, the fact that neuromorphic systems are designed to mimic a [[human brain]] gives rise to unique ethical questions surrounding their usage.<br />
<br />
虽然神经形态工程这个跨学科概念相对较新，但许多适用于'''<font color="#ff8000">类人机器Human-like machines</font>'''和人工智能的伦理考虑大体上也适用于神经形态系统。另一方面，神经形态系统是为了模仿人类大脑而设计，这一事实引起了一些针对性的独特伦理问题。<br />
<br />
However, the practical debate is that neuromorphic hardware as well as artificial "neural networks" are immensely simplified models of how the brain operates or processes information at a much lower [[complex system|complexity in terms of size and functional technology]] and a much more regular structure in terms of [[brain connectivity|connectivity]]. Comparing [[neuromorphic chip]]s to the brain is a very crude comparison similar to comparing a plane to a bird just because they both have wings and a tail. The fact is that neural cognitive systems are many orders of magnitude more [[Energy efficiency (physics)|energy-]] and compute-efficient than current state-of-the-art AI and neuromorphic engineering is an attempt to narrow this gap by inspiring from the brain's mechanism just like many engineering designs have [[bioengineering|bio-inspired features]].<br />
<br />
然而，这种争论的实际情况是，神经形态硬件和人工“神经网络”是对大脑运作或处理信息过程的极其简化的模型，在大小和功能技术方面的'''<font color="#ff8000">复杂性Complexity</font>'''相比而言比较低，在'''<font color="#ff8000">连接性Connectivity</font>'''方面也具有更加规则的结构。将'''<font color="#ff8000">神经形态芯片Neuromorphic chips</font>'''与大脑进行比较是一种非常粗糙的比较，类似于仅仅因为飞机有翅膀和尾巴就将它与鸟进行比较。事实上，当前最先进的人工智能在'''<font color="#ff8000">能耗效率Energy-efficiency</font>'''和计算效率方面距离人脑神经认知系统仍有较大差距，而神经形态工程只是一种通过从大脑机制中获得灵感来缩小这种差距的尝试，就像许多工程设计中都具有'''<font color="#ff8000">生物启发的特征Bio-inspired features</font>'''一样。<br />
<br />
===Democratic concerns===<br />
===公众担忧===<br />
Significant ethical limitations may be placed on neuromorphic engineering due to public perception.<ref name=":27">{{Cite report|url=https://ai100.stanford.edu/sites/g/files/sbiybj9861/f/ai_100_report_0831fnl.pdf|title=Artificial Intelligence and Life in 2030|author=2015 Study Panel|date=September 2016|work=One Hundred Year Study on Artificial Intelligence (AI100)|publisher=Stanford University}}</ref> Special [[Eurobarometer]] 382: Public Attitudes Towards Robots, a survey conducted by the European Commission, found that 60% of [[European Union]] citizens wanted a ban of robots in the care of children, the elderly, or the disabled. Furthermore, 34% were in favor of a ban on robots in education, 27% in healthcare, and 20% in leisure. The European Commission classifies these areas as notably “human.” The report cites increased public concern with robots that are able to mimic or replicate human functions. Neuromorphic engineering, by definition, is designed to replicate the function of the human brain.<ref name=":1">{{Cite web|url=http://ec.europa.eu/commfrontoffice/publicopinion/archives/ebs/ebs_382_en.pdf|title=Special Eurobarometer 382: Public Attitudes Towards Robots|last=European Commission|date=September 2012|website=European Commission}}</ref><br />
<br />
由于公众认知的相关忧虑，神经形态工程学可能会受到严重的伦理限制。<ref name=":27" />欧盟委员会进行的一项调查发现，60% 的欧盟公民希望禁止机器人参与照顾儿童、老人或残疾人的工作。此外，34% 的人支持禁止机器人用于教育，27% 的人支持禁止机器人用于医疗保健，20% 的人支持禁止机器人用于休闲。欧盟委员会将以上领域划入“人类”范畴。报告指出，公众越来越关注能够模仿或复制人类功能的机器人。神经形态工程，顾名思义，是为了复制人脑的功能而设计的。<ref name=":1" /><br />
<br />
The democratic concerns surrounding neuromorphic engineering are likely to become even more profound in the future. The European Commission found that EU citizens between the ages of 15 and 24 are more likely to think of robots as human-like (as opposed to instrument-like) than EU citizens over the age of 55. When presented an image of a robot that had been defined as human-like, 75% of EU citizens aged 15–24 said it corresponded with the idea they had of robots while only 57% of EU citizens over the age of 55 responded the same way. The human-like nature of neuromorphic systems, therefore, could place them in the categories of robots many EU citizens would like to see banned in the future.<ref name=":1" /><br />
<br />
围绕神经形态工程的公众担忧可能在未来变得更加深刻。欧盟委员会发现，相比于55岁以上的欧盟公民，15至24岁的欧盟公民更有可能认为机器人像人(而不是像仪器)。当看到一张“类人”机器人的图片时，年龄在15岁至24岁之间的欧盟公民中有75% 的人表示这符合他们对机器人的想法，而55岁以上的欧盟公民中只有57% 的人有同样的反应。因此，神经形态系统可能因为其类似人类的特性而被归入许多欧盟公民希望在未来禁止使用的机器人类别。<ref name=":1" /><br />
<br />
=== Personhood===<br />
=== 人格问题===<br />
As neuromorphic systems have become increasingly advanced, some scholars{{who|date=August 2021}} have advocated for granting [[personhood]] rights to these systems. If the brain is what grants humans their personhood, to what extent does a neuromorphic system have to mimic the human brain to be granted personhood rights? Critics of technology development in the [[Human Brain Project]], which aims to advance brain-inspired computing, have argued that advancement in neuromorphic computing could lead to machine consciousness or personhood.<ref name=":28">{{Cite journal|last=Aicardi|first=Christine|date=September 2018|title=Accompanying technology development in the Human Brain Project: From foresight to ethics management|journal=Futures|volume=102|pages=114–124|doi=10.1016/j.futures.2018.01.005|doi-access=free}}</ref> If these systems are to be treated as people, critics argue, then many tasks humans perform using neuromorphic systems, including the act of termination of neuromorphic systems, may be morally impermissible as these acts would violate the autonomy of the neuromorphic systems.<ref name=":29">{{Cite journal|last=Lim|first=Daniel|date=2014-06-01|title=Brain simulation and personhood: a concern with the Human Brain Project|journal=Ethics and Information Technology|language=en|volume=16|issue=2|pages=77–89|doi=10.1007/s10676-013-9330-5|s2cid=17415814|issn=1572-8439}}</ref><br />
<br />
随着神经形态系统的日益发展，一些学者主张赋予这些系统'''<font color="ff8000">人格Personhood</font>'''权。如果是大脑赋予了人类人格，那么在多大程度上模仿人类大脑的神经形态系统才能被赋予人格权利？“人类大脑计划”旨在推进以大脑为灵感的计算机技术发展，该计划的批评者认为，神经形态计算机技术的进步可能导致机器意识或人格的形成。<ref name=":28" />这些批评者认为，如果这些系统被当作人来对待，那么人类使用神经形态系统执行任务（包括终止神经形态系统）的行为，在道德上就可能是不被允许的，因为这些行为将违反神经形态系统的自主性。<ref name=":29" /><br />
<br />
<br />
==Dual use (military applications)==<br />
==军民两用技术（军事应用）==<br />
The [[Joint Artificial Intelligence Center]], a branch of the U.S. military, is a center dedicated to the procurement and implementation of AI software and neuromorphic hardware for combat use. Specific applications include smart headsets/goggles and robots. JAIC intends to rely heavily on neuromorphic technology to connect "every fighter every shooter" within a network of neuromorphic-enabled units.<br />
<br />
'''<font color="#ff8000">联合人工智能中心The Joint Artificial Intelligence Center</font>'''(JAIC),是美国军队的一个分支，专门从事采购和实施用于战斗的人工智能软件和神经形态硬件。具体应用包括智能耳机/护目镜和机器人。JAIC打算高度依赖神经形态技术，'''<font color="#32CD32">使用神经形态技术来连接神经形态单位网络中的“每个战士每个射手”</font>'''。<br />
<br />
<br />
==Legal considerations==<br />
==法律问题==<br />
Skeptics have argued that there is no way to apply the electronic personhood, the concept of personhood that would apply to neuromorphic technology, legally. In a letter signed by 285 experts in law, robotics, medicine, and ethics opposing a European Commission proposal to recognize “smart robots” as legal persons, the authors write, “A legal status for a robot can’t derive from the [[Natural person|Natural Person]] model, since the robot would then hold [[human rights]], such as the right to dignity, the right to its integrity, the right to remuneration or the right to citizenship, thus directly confronting the Human rights. This would be in contradiction with the [[Charter of Fundamental Rights of the European Union]] and the [[Convention for the Protection of Human Rights and Fundamental Freedoms]].”<ref name=":30">{{Cite web|url=http://www.robotics-openletter.eu/|title=Robotics Openletter {{!}} Open letter to the European Commission|language=fr-FR|access-date=2019-05-10}}</ref><br />
<br />
质疑者者认为，在法律上没有办法应用能够适用于神经形态技术的电子人格。在一封由285名法律、机器人技术、医学和伦理学专家的联名信中，作者们反对欧盟委员会提出的承认“智能机器人”为法人的提议。他们写道，“机器人的法律地位无法从'''<font color="#ff8000">自然人Natural Person</font>'''模型中推导出来，因为机器人将被赋予'''<font color="#ff8000">人权Human rights</font>'''，如尊严权、完整权、报酬权或公民权，从而直接面临人权问题。这将有悖于《欧联基本权利宪章》和《欧洲保障人权和根本自由公约》”。<ref name=":30" /><br />
<br />
===Ownership and property rights ===<br />
===所有权及财产权问题===<br />
There is significant legal debate around property rights and artificial intelligence. In ''Acohs Pty Ltd v. Ucorp Pty Ltd'', Justice Christopher Jessup of the [[Federal Court of Australia]] found that the [[source code]] for [[Material safety data sheets|Material Safety Data Sheets]] could not be [[Copyright law of Australia|copyrighted]] as it was generated by a [[software interface]] rather than a human author.<ref name=":31">{{Cite web|url=http://www.lavan.com.au/advice/intellectual_property_technology/copyright_in_source_code_and_digital_products|title=Copyright in source code and digital products|last=Lavan|website=Lavan|language=en|access-date=2019-05-10}}</ref> The same question may apply to neuromorphic systems: if a neuromorphic system successfully mimics a human brain and produces a piece of original work, who, if anyone, should be able to claim ownership of the work?<ref name=":32">{{cite journal |last1=Eshraghian|first1=Jason K. |title=Human Ownership of Artificial Creativity |journal=Nature Machine Intelligence |date=9 March 2020 |volume=2 |pages=157–160 |doi=10.1038/s42256-020-0161-x}}</ref><br />
<br />
法律界围绕财产权和人工智能有着重大争论。在Acohs Pty有限公司诉Ucorp Pty有限公司一案中，澳大利亚联邦法院的克里斯托弗·杰瑟普法官发现，版权保护不适用于材料安全数据表的'''<font color="#ff8000">源代码Source code</font>'''，因为它是由'''<font color="#ff8000">软件界面Software interface</font>'''生成而非人类工作者生成的。<ref name=":31" />同样的问题可能也适用于神经形态系统: 如果一个神经形态系统成功地模仿了人类的大脑，并产生了一部原创作品，那么该如何确认这部作品的所有权归属？<ref name=":32" /><br />
<br />
<br />
==Neuromemristive systems==<br />
==神经记忆电阻系统==<br />
Neuromemristive systems are a subclass of neuromorphic computing systems that focus on the use of [[memristors]] to implement [[neuroplasticity]]. While neuromorphic engineering focuses on mimicking biological behavior, neuromemristive systems focus on abstraction.<ref name=":33">{{Cite web|url=https://digitalops.sandia.gov/Mediasite/Play/a10cf6ceb55d47608bb8326dd00e46611d|title=002.08 N.I.C.E. Workshop 2014: Towards Intelligent Computing with Neuromemristive Circuits and Systems - Feb. 2014|website=digitalops.sandia.gov|access-date=2019-08-26}}</ref> For example, a neuromemristive system may replace the details of a [[Cerebral cortex|cortical]] microcircuit's behavior with an abstract neural network model.<ref name=":34">C. Merkel and D. Kudithipudi, "Neuromemristive extreme learning machines for pattern classification," ISVLSI, 2014.</ref><br />
<br />
神经记忆电阻系统是神经形态计算系统的一个亚类，主要研究利用'''<font color="ff8000">记忆电阻器Memristors</font>'''实现'''<font color="ff8000">神经可塑性Neuroplasticity</font>'''。神经形态工程的重点是模拟生物行为，而神经记忆电阻系统的重点是提取。<ref name=":33" />举个例子，一个神经记忆系统可能用抽象的神经网络模型替代'''<font color="ff8000">皮层Cortical</font>'''微电路的行为细节。<ref name=":34" /><br />
<br />
There exist several neuron inspired threshold logic functions<ref name="Maan 1–13" /> implemented with memristors that have applications in high level [[pattern recognition]] applications. Some of the applications reported recently include [[speech recognition]],<ref name=":35">{{Cite journal|title = Memristor pattern recogniser: isolated speech word recognition|journal = Electronics Letters|pages = 1370–1372|volume = 51|issue = 17|doi = 10.1049/el.2015.1428|first1 = A.K.|last1 = Maan|first2 = A.P.|last2 = James|first3 = S.|last3 = Dimitrijev|year = 2015|bibcode = 2015ElL....51.1370M|hdl = 10072/140989|s2cid = 61454815|url = https://semanticscholar.org/paper/48d3ab11ec6e213b62f11eedcfb7b7febb058674|hdl-access = free}}</ref> [[face recognition]]<ref name=":36">{{Cite journal|title = Memristive Threshold Logic Face Recognition|journal = Procedia Computer Science|date = 2014-01-01|pages = 98–103|volume = 41|series = 5th Annual International Conference on Biologically Inspired Cognitive Architectures, 2014 BICA|doi = 10.1016/j.procs.2014.11.090|first1 = Akshay Kumar|last1 = Maan|first2 = Dinesh S.|last2 = Kumar|first3 = Alex Pappachen|last3 = James|doi-access = free}}</ref> and [[object recognition]].<ref name=":37">{{Cite journal|title = Memristive Threshold Logic Circuit Design of Fast Moving Object Detection|journal = IEEE Transactions on Very Large Scale Integration (VLSI) Systems|date = 2015-10-01|issn = 1063-8210|pages = 2337–2341|volume = 23|issue = 10|doi = 10.1109/TVLSI.2014.2359801|first1 = A.K.|last1 = Maan|first2 = D.S.|last2 = Kumar|first3 = S.|last3 = Sugathan|first4 = A.P.|last4 = James|arxiv = 1410.1267|s2cid = 9647290}}</ref> They also find applications in replacing conventional digital logic gates.<ref name=":38">{{Cite journal|title = Resistive Threshold Logic|journal = IEEE Transactions on Very Large Scale Integration (VLSI) Systems|date = 2014-01-01|issn = 1063-8210|pages = 190–195|volume = 22|issue = 1|doi = 10.1109/TVLSI.2012.2232946|first1 = A.P.|last1 = James|first2 = L.R.V.J.|last2 = Francis|first3 = D.S.|last3 = Kumar|arxiv = 1308.0090|s2cid = 7357110}}</ref><ref name=":39">{{Cite journal|title = Threshold Logic Computing: Memristive-CMOS Circuits for Fast Fourier Transform and Vedic Multiplication|journal = IEEE Transactions on Very Large Scale Integration (VLSI) Systems|date = 2015-11-01|issn = 1063-8210|pages = 2690–2694|volume = 23|issue = 11|doi = 10.1109/TVLSI.2014.2371857|first1 = A.P.|last1 = James|first2 = D.S.|last2 = Kumar|first3 = A.|last3 = Ajayan|arxiv = 1411.5255|s2cid = 6076956}}</ref><br />
<br />
受神经元启发、使用记忆电阻器实现的阈值逻辑函数<ref name="Maan 1–13" />在高级'''<font color="#ff8000">模式识别Pattern recognition</font>'''中有着广泛的应用，最近报道中其应用包括'''<font color="#ff8000">语音识别Speech recognition<ref name=":35" /></font>'''、'''<font color="#ff8000">人脸识别Face recognition<ref name=":36" /></font>'''和'''<font color="#ff8000">物体识别Object recognition<ref name=":37" /></font>'''。阈值逻辑函数还可以用来取代传统的数字逻辑门。<ref name=":38" /><ref name=":39" /><br />
<br />
For ideal passive memristive circuits there is an exact equation (Caravelli-Traversa-[[Di Ventra]] equation) for the internal memory of the circuit:<ref name=":40">{{cite journal |last=Caravelli |display-authors=etal|arxiv=1608.08651 |title=The complex dynamics of memristive circuits: analytical results and universal slow relaxation |year=2017 |doi=10.1103/PhysRevE.95.022140 |pmid= 28297937 |volume=95 |issue= 2 |pages= 022140 |journal=Physical Review E|bibcode=2017PhRvE..95b2140C |s2cid=6758362}}</ref><br />
<br />
对于理想的无源记忆电路，电路的内部记忆可以用精确的方程(Caravelli-Traversa-Di Ventra方程) 来描述:<ref name=":40" /><br />
<br />
:<math> \frac{d}{dt} \vec{W} = \alpha \vec{W}-\frac{1}{\beta} (I+\xi \Omega W)^{-1} \Omega \vec S </math><br />
<br />
as a function of the properties of the physical memristive network and the external sources. In the equation above, <math>\alpha</math> is the "forgetting" time scale constant, <math> \xi=r-1</math> and <math>r=\frac{R_\text{off}}{R_\text{on}}</math> is the ratio of ''off'' and ''on'' values of the limit resistances of the memristors, <math> \vec S </math> is the vector of the sources of the circuit and <math>\Omega</math> is a projector on the fundamental loops of the circuit. The constant <math>\beta</math> has the dimension of a voltage and is associated to the properties of the [[memristor]]; its physical origin is the charge mobility in the conductor. The diagonal matrix and vector <math>W=\operatorname{diag}(\vec W)</math> and <math>\vec W<\math> respectively, are instead the internal value of the memristors, with values between 0 and 1. This equation thus requires adding extra constraints on the memory values in order to be reliable.<br />
<br />
Caravelli-Traversa-Di Ventra方程是描述物理记忆网络和外部源性质的函数。在上述方程中，<math>\alpha</math>是“遗忘”时间尺度常数，<math>\xi=r-1</math>，<math>r =\frac{R\text_{off}}{R_\text{on}}</math>是记忆电阻器off状态和on状态极限电阻值之比，<math>\vec S</math>是电路源的矢量，<math>\Omega</math>是电路基本环路的投影。常数<math>\beta</math>具有电压的量纲，与记忆电阻器的特性有关；其物理原型是导体中的电荷迁移率。对角矩阵和向量 <math>W=\operatorname{diag}(\vec W)</math>和<math>\vec W</math> '''<font color="#32CD32">分别是记忆电阻器的内阻</font>'''，值在0到1之间。因此，这个等式需要在'''<font color="32CD32">内存值</font>'''上添加额外约束以保证可靠性。<br />
<br />
<br />
==See also==<br />
==相关词条==<br />
{{Columns-list|colwidth=18em|<br />
* [[AI accelerator (computer hardware)|AI accelerator]]<br />
* [[Artificial brain]]<br />
* [[Biomorphic]]<br />
* [[Cognitive computer]]<br />
* [[Computation and Neural Systems]]<br />
* [[Differentiable programming]]<br />
* [[Event camera]]<br />
* [[Neurorobotics]]<br />
* [[Optical flow sensor]]<br />
* [[Physical neural network]]<br />
* [[SpiNNaker]]<br />
* [[SyNAPSE]]<br />
* [[Retinomorphic sensor]]<br />
* [[Vision chip]]<br />
* [[Vision processing unit]]<br />
* [[Zeroth (software)]]<br />
* [[Hardware for artificial intelligence]]<br />
}}<br />
<br />
{{Portal bar|Electronics}}<br />
<br />
==References==<br />
{{Reflist|40em}}<br />
<br />
==External links==<br />
<!--======================== {{No more links}} ============================<br />
| PLEASE BE CAUTIOUS IN ADDING MORE LINKS TO THIS ARTICLE. Wikipedia |<br />
| is not a collection of links nor should it be used for advertising. |<br />
| |<br />
| Excessive or inappropriate links WILL BE DELETED. |<br />
| See [[Wikipedia:External links]] & [[Wikipedia:Spam]] for details. |<br />
| |<br />
| If there are already plentiful links, please propose additions or |<br />
| replacements on this article's discussion page, or submit your link |<br />
| to the relevant category at the Open Directory Project (dmoz.org) |<br />
| and link back to that category using the {{dmoz}} template. |<br />
======================={{No more links}}=============================--><br />
{{Commons category|Neuromorphic Engineering}}<br />
*[https://web.archive.org/web/20150727034331/http://ine-web.org/workshops/workshops-overview Telluride Neuromorphic Engineering Workshop]<br />
*[https://archive.today/20130115190057/http://capocaccia.ethz.ch/ CapoCaccia Cognitive Neuromorphic Engineering Workshop]<br />
*[https://web.archive.org/web/20190716132350/http://www.ine-web.org/ Institute of Neuromorphic Engineering]<br />
*[http://www.ine-news.org/ INE news site].<br />
*[http://www.frontiersin.org/neuromorphic_engineering Frontiers in Neuromorphic Engineering Journal]<br />
*[http://www.cns.caltech.edu/ Computation and Neural Systems] department at the [[California Institute of Technology]].<br />
*[http://www.humanbrainproject.eu/ Human Brain Project official site]<br />
*[https://www.the-scientist.com/features/building-a-silicon-brain-65738 Building a Silicon Brain:] Computer chips based on biological neurons may help simulate larger and more-complex brain models. May 1, 2019. SANDEEP RAVINDRAN<br />
<br />
<br />
*Telluride Neuromorphic Engineering Workshop<br />
*CapoCaccia Cognitive Neuromorphic Engineering Workshop<br />
*Institute of Neuromorphic Engineering<br />
*INE news site.<br />
*Frontiers in Neuromorphic Engineering Journal<br />
*Computation and Neural Systems department at the California Institute of Technology.<br />
*Human Brain Project official site<br />
*Building a Silicon Brain: Computer chips based on biological neurons may help simulate larger and more-complex brain models. May 1, 2019. SANDEEP RAVINDRAN<br />
<br />
=<nowiki>外部链接</nowiki>= <br />
*碲化物神经形态工程工作室<br />
*CapoCaccia认知神经形态工程工作室<br />
*神经形态工程研究所<br />
*INE新闻站点。<br />
*《神经形态工程学前沿》<br />
*加州理工学院计算与神经系统系。<br />
*人脑项目官方网站<br />
*打造硅脑: 基于生物神经元的计算机芯片可能有助于模拟更大、更复杂的大脑模型。2019年5月1日。SANDEEP RAVINDRAN.<br />
<br />
{{Differentiable computing}}<br />
{{Authority control}}<br />
<br />
[[Category:Electrical engineering]]<br />
[[Category:Neuroscience]]<br />
[[分类:AI accelerators|*]]<br />
[[Category:Artificial intelligence]]<br />
[[Category:Robotics]]<br />
<br />
Category:Electrical engineering<br />
Category:Neuroscience<br />
*<br />
Category:Artificial intelligence<br />
Category:Robotics<br />
<br />
类别: 电气工程类别: 神经科学 * 类别: 人工智能类别: 机器人<br />
<br />
<noinclude><br />
<br />
<small>This page was moved from [[wikipedia:en:Neuromorphic engineering]]. Its edit history can be viewed at [[神经形态计算/edithistory]]</small></noinclude><br />
<br />
[[Category:待整理页面]]<br />
<br />
<references /></div>Shenky20https://wiki.swarma.org/index.php?title=%E7%A5%9E%E7%BB%8F%E5%BD%A2%E6%80%81%E8%AE%A1%E7%AE%97&diff=30752神经形态计算2022-04-28T11:55:53Z<p>Shenky20：/* 神经记忆电阻系统 */</p>
<hr />
<div><br />
此词条由神经动力学模型读书会词条梳理志愿者Shenky20翻译审校，未经专家审校，带来阅读不便，请见谅。<br />
<br />
{{Use American English|date = January 2019}}<br />
{{Use mdy dates|date = January 2019}}<br />
<br />
'''Neuromorphic engineering''', also known as '''neuromorphic computing''',<ref name=":3">{{Cite journal | doi = 10.1145/2601069| title = Neuromorphic computing gets ready for the (really) big time| journal = [[Communications of the ACM]]| volume = 57| issue = 6| year = 2014| last1 = Monroe | first1 = D. | pages = 13–15| s2cid = 20051102}}</ref><ref name=":4">{{Cite journal | doi = 10.1088/0957-4484/21/17/175202| pmid = 20368686| title = Nanotube devices based crossbar architecture: Toward neuromorphic computing| journal = Nanotechnology| volume = 21| issue = 17| pages = 175202| year = 2010| last1 = Zhao | first1 = W. S. | last2 = Agnus | first2 = G. | last3 = Derycke | first3 = V. | last4 = Filoramo | first4 = A. | last5 = Bourgoin | first5 = J. -P. | last6 = Gamrat | first6 = C. | bibcode = 2010Nanot..21q5202Z| url = https://zenodo.org/record/3428659}}</ref><ref name="humanbrainproject">{{YouTube|id=6RoiZ90mGfw|title=The Human Brain Project SP 9: Neuromorphic Computing Platform}}</ref> is the use of [[very-large-scale integration]] (VLSI) systems containing electronic [[analog circuit]]s to mimic neuro-biological architectures present in the nervous system. A neuromorphic computer/chip is any device that uses physical artificial neurons (made from silicon) to do computations.<ref name=":5">{{cite journal|last1=Mead|first1=Carver|title=Neuromorphic electronic systems|journal=Proceedings of the IEEE|date=1990|volume=78|issue=10|pages=1629–1636|doi=10.1109/5.58356|url=https://authors.library.caltech.edu/53090/1/00058356.pdf}}</ref><ref name=":2" /> In recent times, the term ''neuromorphic'' has been used to describe analog, digital, mixed-mode analog/digital VLSI, and software systems that implement models of [[neural system]]s (for [[perception]], [[motor control]], or [[multisensory integration]]). The implementation of neuromorphic computing on the hardware level can be realized by oxide-based [[memristor]]s,<ref name="Maan 1–13">{{Cite journal|last1=Maan|first1=A. K.|last2=Jayadevi|first2=D. A.|last3=James|first3=A. P.|date=2016-01-01|title=A Survey of Memristive Threshold Logic Circuits|journal=IEEE Transactions on Neural Networks and Learning Systems|volume=PP|issue=99|pages=1734–1746|doi=10.1109/TNNLS.2016.2547842|pmid=27164608|issn=2162-237X|arxiv=1604.07121|bibcode=2016arXiv160407121M|s2cid=1798273}}</ref> spintronic memories, threshold switches, and [[transistor]]s.<ref name=":6">{{Cite journal|title = Mott Memory and Neuromorphic Devices|journal = Proceedings of the IEEE|date = 2015-08-01|issn = 0018-9219|pages = 1289–1310|volume = 103|issue = 8|doi = 10.1109/JPROC.2015.2431914|first1 = You|last1 = Zhou|first2 = S.|last2 = Ramanathan|s2cid = 11347598|url = https://zenodo.org/record/895565}}</ref><ref name=":2">{{Cite document|title=Neuromorphic Circuits With Neural Modulation Enhancing the Information Content of Neural Signaling {{!}} International Conference on Neuromorphic Systems 2020|language=EN|doi=10.1145/3407197.3407204|s2cid=220794387}}</ref> Training software-based neuromorphic systems of spiking neural networks can be achieved using error backpropagation, e.g., using [[Python (programming language)|Python]] based frameworks such as snnTorch,<ref name=":7">{{cite journal |last1=Eshraghian|first1=Jason K.|last2=Ward|first2=Max|last3=Neftci |first3=Emre|last4=Wang|first4=Xinxin|last5=Lenz|first5=Gregor|last6=Dwivedi|first6=Girish|last7=Bennamoun|first7=Mohammed|last8=Jeong|first8=Doo Seok|last9=Lu|first9=Wei D.|title=Training Spiking Neural Networks Using Lessons from Deep Learning |date=1 October 2021 |arxiv=2109.12894 }}</ref> or using canonical learning rules from the biological learning literature, e.g., using BindsNet.<ref name=":8">{{Cite web | url=https://github.com/Hananel-Hazan/bindsnet | title=Hananel-Hazan/bindsnet: Simulation of spiking neural networks (SNNs) using PyTorch.| date=31 March 2020}}</ref><br />
<br />
<br />
'''<font color="#ff8000">神经形态工程Neuromorphic engineering</font>'''(也称为'''<font color="#ff8000">神经形态计算Neuromorphic computing</font>''')<ref name=":3" /><ref name=":4" /><ref name="humanbrainproject" />是指使用包含电子'''<font color="#ff8000">模拟电路Analog circuit</font>'''的'''<font color="#ff8000">超大规模集成电路Very-large-scale integration</font>'''系统来模拟神经系统中生理结构的研究方法。神经形态计算机或神经形态芯片包括任何使用由硅制成的人造神经元进行计算的设备。<ref name=":5" /><ref name=":2" />近年来，神经形态学（neuromorphic）这个术语被用来描述能够实现'''<font color="#ff8000">神经系统Neural system</font>'''模型功能(如'''<font color="#ff8000">感知Perception</font>'''、'''<font color="#ff8000">运动控制Motor control</font>'''，'''<font color="#ff8000">多感官整合Multisensory integration</font>'''等)的模拟、数字、模拟/数字混合模式超大规模集成电路和软件系统。神经形态计算的硬件实现可以通过基于氧化物的'''<font color="#ff8000">记忆电阻器Memristor</font>'''、<ref name="Maan 1–13" />自旋电子存储器、阈值开关和'''<font color="#ff8000">晶体管Transistor</font>'''来实现。<ref name=":6" /><ref name=":2" />对基于软件的脉冲神经网络系统的训练可以通过误差反向传播机制来实现，例如，使用snnTorch等基于Python的框架，<ref name=":7" />或使用BindsNet等典型的受生物启发的学习模式。<ref name=":8" /><br />
<br />
A key aspect of neuromorphic engineering is understanding how the morphology of individual neurons, circuits, applications, and overall architectures creates desirable computations, affects how information is represented, influences robustness to damage, incorporates learning and development, adapts to local change (plasticity), and facilitates evolutionary change.<br />
<br />
<br />
神经形态工程领域的一个关键问题，就是理解'''<font color="#32CD32">单个神经元形态、神经回路、应用和整体结构</font>'''如何产生理想的计算，如何影响信息的表示和对破坏的鲁棒性，如何整合学习和发展，如何适应局部变化(可塑性) 并促进逐渐发展的变化。<br />
<br />
Neuromorphic engineering is an interdisciplinary subject that takes inspiration from [[biology]], [[physics]], [[mathematics]], [[computer science]], and [[electronic engineering]]<ref name=":2" /> to design artificial neural systems, such as [[Machine vision|vision systems]], head-eye systems, auditory processors, and autonomous robots, whose physical architecture and design principles are based on those of biological nervous systems.<ref name=":9">{{Cite journal | doi = 10.1155/2012/705483| title = Qualitative Functional Decomposition Analysis of Evolved Neuromorphic Flight Controllers| journal = Applied Computational Intelligence and Soft Computing| volume = 2012| pages = 1–21| year = 2012| last1 = Boddhu | first1 = S. K. | last2 = Gallagher | first2 = J. C. | doi-access = free}}</ref> It was developed by [[Carver Mead]]<ref name=":10">{{cite web|title=carver mead website|last1=Mead|first1=Carver|url=http://www.carvermead.caltech.edu/index.html|website=carvermead}}</ref> in the late 1980s.<br />
<br />
<br />
神经形态工程是以'''<font color="#ff8000">生物学Biology</font>'''、'''<font color="#ff8000">物理学Physics</font>'''、'''<font color="#ff8000">数学Mathematics</font>'''、'''<font color="#ff8000">计算机科学Computer science</font>'''和'''<font color="#ff8000">电子工程Electronic engineering<ref name=":2" /></font>'''等学科为基础，设计人工神经系统（如'''<font color="#ff8000">视觉系统Vision systems</font>'''、头眼系统、听觉处理器以及物理结构和设计原则都受启发于生物神经系统的自主机器人）的一门交叉学科。<ref name=":9" />20世纪80年代后期，卡弗·米德极大地推动了神经形态工程领域的发展。<ref name=":10" /><br />
<br />
==Neurological inspiration==<br />
==神经生物学启发==<br />
Neuromorphic engineering is set apart by the inspiration it takes from what we know about the structure and operations of the [[brain]]. Neuromorphic engineering translates what we know about the brain's function into computer systems. Work has mostly focused on replicating the analog nature of [[biological computation]] and the role of [[neuron]]s in [[cognition]].<br />
<br />
<br />
神经形态工程形成的灵感来源于现已知的'''<font color="#ff8000">大脑Brain</font>'''结构及运作机制知识，它将我们对大脑功能的了解用于研究和优化计算机系统。该领域工作主要集中于对'''<font color="#ff8000">生物计算Biological computation</font>'''模拟特性和'''<font color="#ff8000">神经元Neuron</font>'''在'''<font color="#ff8000">认知Cognition</font>'''中所发挥作用的复刻。<br />
<br />
The biological processes of neurons and their [[synapse]]s are dauntingly complex, and thus very difficult to artificially simulate. A key feature of biological brains is that all of the processing in neurons use analog chemical signals. This makes it hard to replicate brains in computers because the current generation of computers is completely digital. However, the characteristics of these parts can be abstracted into mathematical functions that closely capture the essence of the neuron's operations.<br />
<br />
神经元及其'''<font color="#ff8000">突触Synapse</font>'''的生理学过程极其复杂，难以进行人工模拟。神经元在所有处理过程中都使用化学模拟信号，这是大脑的一个关键生理学特征。这个特征大大增加了在计算机中复制大脑的难度，因为目前的计算机是完全数字化的。然而，部分特征可以抽象为数学函数，这些函数能够紧密捕捉神经元运行的本质。<br />
<br />
The goal of neuromorphic computing is not to perfectly mimic the brain and all of its functions, but instead to extract what is known of its structure and operations to be used in a practical computing system. No neuromorphic system will claim nor attempt to reproduce every element of neurons and synapses, but all adhere to the idea that computation is highly [[distributed processing|distributed]] throughout a series of small computing elements analogous to a neuron. While this sentiment is standard, researchers chase this goal with different methods.<ref name=":11">{{Cite journal | doi = 10.1088/1741-2560/13/5/051001| title = Large-scale neuromorphic computing systems| journal = Journal of Neural Engineering| volume = 13| pages = 1–15| year = 2016| last1 = Furber | first1 = Steve| issue = 5| pmid = 27529195| bibcode = 2016JNEng..13e1001F| doi-access = free}}</ref><br />
<br />
神经形态计算的目标不是完美地模拟大脑及其所有功能，而是利用已知的大脑结构和运转机制来研发或优化实际的计算系统。任何神经形态学系统都不会声称或试图复制神经元和突触中的每一个元素，但所有人都一致认可将计算高度'''<font color="#ff8000">分散Distribute</font>'''于一系列类似于神经元的小型计算元素的理念。研究人员用不同的方法来追求这一普遍目标。<ref name=":11" /><br />
<br />
<br />
==Examples==<br />
==代表性成果==<br />
As early as 2006, researchers at [[Georgia Tech]] published a field programmable neural array.<ref name=":12">{{Cite book|title = A field programmable neural array|last1 = Farquhar|first1 = Ethan|date = May 2006|journal = IEEE International Symposium on Circuits and Systems|pages = 4114–4117|last2 = Hasler|first2 = Paul.|doi = 10.1109/ISCAS.2006.1693534|isbn = 978-0-7803-9389-9|s2cid = 206966013}}</ref> This chip was the first in a line of increasingly complex arrays of floating gate transistors that allowed programmability of charge on the gates of [[MOSFET]]s to model the channel-ion characteristics of neurons in the brain and was one of the first cases of a silicon programmable array of neurons.<br />
<br />
早在2006年，佐治亚理工学院的研究人员就研发出了一种现场可编程神经阵列。<ref name=":12" />在此之后，一系列越来越复杂的浮栅晶体管阵列被成功研发出来，这些晶体管阵列可以通过在'''<font color="#ff8000">金属-氧化物半导体效应晶体管MOSFET</font>'''的栅极上编程来模拟大脑中神经元的离子通道特性，这也是以硅为主要材料的可编程神经元阵列的最早成功案例之一。<br />
<br />
In November 2011, a group of [[MIT]] researchers created a computer chip that mimics the analog, ion-based communication in a synapse between two neurons using 400 transistors and standard [[CMOS]] manufacturing techniques.<ref name=":13">{{cite web|title=MIT creates "brain chip"|url=http://www.extremetech.com/extreme/105067-mit-creates-brain-chip|access-date=4 December 2012}}</ref><ref name="Neuromorphic silicon paper">{{cite journal|title=Neuromorphic silicon neurons and large-scale neural networks: challenges and opportunities|doi=10.3389/fnins.2011.00108|pmid=21991244|pmc=3181466|volume=5|pages=108|journal=Frontiers in Neuroscience|year=2011|last1=Poon|first1=Chi-Sang|last2=Zhou|first2=Kuan|doi-access=free}}</ref><br />
<br />
2011年11月，麻省理工学院的一组研究人员研发出一种计算机芯片，该芯片上使用标准的'''<font color="#ff8000">互补金属氧化物半导体CMOS</font>'''制造技术集成了400个晶体管，能够模拟神经元间突触中基于离子的通讯。<ref name=":13" /><ref name="Neuromorphic silicon paper" /><br />
<br />
In June 2012, [[spintronic]] researchers at [[Purdue University]] presented a paper on the design of a neuromorphic chip using [[Spin valve|lateral spin valve]]s and [[memristor]]s. They argue that the architecture works similarly to neurons and can therefore be used to test methods of reproducing the brain's processing. In addition, these chips are significantly more energy-efficient than conventional ones.<ref name="Spin Devices Prop">{{Cite arXiv|title=Proposal For Neuromorphic Hardware Using Spin Devices|eprint=1206.3227|last1=Sharad|first1=Mrigank|last2=Augustine|first2=Charles|last3=Panagopoulos|first3=Georgios|last4=Roy|first4=Kaushik|class=cond-mat.dis-nn|year=2012}}</ref><br />
<br />
2012年6月，普渡大学的'''<font color="#ff8000">自旋电子学Spintronic</font>'''研究人员发表了一篇关于利用'''<font color="#ff8000">侧向自旋阀Lateral spin valves</font>'''和'''<font color="#ff8000">记忆电阻器Memristors</font>'''设计神经形态芯片的论文。他们认为，这种芯片结构的工作原理与神经元相似，因此可以用于大脑运行机制的复刻方法的测试。此外，这些芯片在能耗方面明显优于传统芯片。<ref name="Spin Devices Prop" /><br />
<br />
Research at [[HP Labs]] on Mott memristors has shown that while they can be non-[[Volatile memory|volatile]], the volatile behavior exhibited at temperatures significantly below the [[phase transition]] temperature can be exploited to fabricate a [[neuristor]],<ref name=":0" /> a biologically-inspired device that mimics behavior found in neurons.<ref name=":0">{{Cite journal | doi = 10.1038/nmat3510| pmid = 23241533| title = A scalable neuristor built with Mott memristors| journal = Nature Materials| volume = 12| issue = 2| pages = 114–7| year = 2012| last1 = Pickett | first1 = M. D. | last2 = Medeiros-Ribeiro | first2 = G. | last3 = Williams | first3 = R. S. | bibcode = 2013NatMa..12..114P| s2cid = 16271627| url = https://semanticscholar.org/paper/b6ba6f496ace2b947f111059663e76bb60e9efeb}}</ref> In September 2013, they presented models and simulations that show how the spiking behavior of these neuristors can be used to form the components required for a [[Turing machine]].<ref name=":14">{{cite journal|doi=10.1088/0957-4484/24/38/384002|title=Phase transitions enable computational universality in neuristor-based cellular automata|author1=Matthew D Pickett|author2=R Stanley Williams|name-list-style=amp|date=September 2013|publisher=IOP Publishing Ltd|journal=Nanotechnology|volume=24|issue=38|pmid=23999059|bibcode=2013Nanot..24L4002P|at=384002}}</ref><br />
<br />
'''<font color="#ff8000">惠普实验室HP labs</font>'''在莫特记忆电阻器上的研究表明，尽管它们可以是非'''<font color="#ff8000">易失性Volatile</font>'''的，但是在'''<font color="#ff8000">相变Phase transition</font>'''温度以下时表现出的易失性行为可以被用来制造'''<font color="#ff8000">类神经元电阻器Neuristor</font>'''(一种生物学启发的模仿神经元行为的硬件)<ref name=":0" />。2013年9月，他们通过模型和仿真展示了这些类神经元电阻器的脉冲行为如何产生'''<font color="#ff8000">图灵机Turing machine</font>'''的所需元素。<ref name=":14" /><br />
<br />
[[Neurogrid]], built by ''Brains in Silicon'' at [[Stanford University]],<ref name=":15">{{cite journal|last1=Boahen|first1=Kwabena|title=Neurogrid: A Mixed-Analog-Digital Multichip System for Large-Scale Neural Simulations|journal=Proceedings of the IEEE|date=24 April 2014|volume=102|issue=5|pages=699–716|doi=10.1109/JPROC.2014.2313565|s2cid=17176371}}</ref> is an example of hardware designed using neuromorphic engineering principles. The circuit board is composed of 16 custom-designed chips, referred to as NeuroCores. Each NeuroCore's analog circuitry is designed to emulate neural elements for 65536 neurons, maximizing energy efficiency. The emulated neurons are connected using digital circuitry designed to maximize spiking throughput.<ref name=":16">{{cite journal|doi=10.1038/503022a|pmid = 24201264|title = Neuroelectronics: Smart connections|journal = Nature|volume = 503|issue = 7474|pages = 22–4|year = 2013|last1 = Waldrop|first1 = M. Mitchell|bibcode = 2013Natur.503...22W|doi-access = free}}</ref><ref name=":17">{{cite journal|doi=10.1109/JPROC.2014.2313565|title = Neurogrid: A Mixed-Analog-Digital Multichip System for Large-Scale Neural Simulations|journal = Proceedings of the IEEE|volume = 102|issue = 5|pages = 699–716|year = 2014|last1 = Benjamin|first1 = Ben Varkey|last2 = Peiran Gao|last3 = McQuinn|first3 = Emmett|last4 = Choudhary|first4 = Swadesh|last5 = Chandrasekaran|first5 = Anand R.|last6 = Bussat|first6 = Jean-Marie|last7 = Alvarez-Icaza|first7 = Rodrigo|last8 = Arthur|first8 = John V.|last9 = Merolla|first9 = Paul A.|last10 = Boahen|first10 = Kwabena|s2cid = 17176371}}</ref><br />
<br />
'''<font color="#ff8000">神经栅格Neurogrid<ref name=":15" /></font>'''是由斯坦福大学Brains in Silicon公司研发的、使用神经形态工程原理设计的硬件。该电路板由16个定制设计的芯片组成(NeuroCores)。在设计中，每个NeuroCore芯片的模拟电路对65536个神经元的神经元素进行模拟，以最大限度地提高能量效率。模拟出的神经元通过设计的数字电路连接，以最大化脉冲吞吐量。<ref name=":16" /><ref name=":17" /><br />
<br />
A research project with implications for neuromorphic engineering is the [[Human Brain Project]] that is attempting to simulate a complete human brain in a supercomputer using biological data. It is made up of a group of researchers in neuroscience, medicine, and computing.<ref name=":18">{{cite web|title=Involved Organizations|url=http://www.humanbrainproject.eu/partners.html|access-date=22 February 2013|url-status=dead|archive-url=https://web.archive.org/web/20130302142627/http://www.humanbrainproject.eu/partners.html|archive-date=2 March 2013}}</ref> [[Henry Markram]], the project's co-director, has stated that the project proposes to establish a foundation to explore and understand the brain and its diseases, and to use that knowledge to build new computing technologies. The three primary goals of the project are to better understand how the pieces of the brain fit and work together, to understand how to objectively diagnose and treat brain diseases, and to use the understanding of the human brain to develop neuromorphic computers. That the simulation of a complete human brain will require a supercomputer a thousand times more powerful than today's encourages the current focus on neuromorphic computers.<ref name=":19">{{cite web|title=Human Brain Project|url=http://www.humanbrainproject.eu|access-date=22 February 2013}}</ref> $1.3 billion has been allocated to the project by The [[European Commission]].<ref name=":20">{{cite web|title=The Human Brain Project and Recruiting More Cyberwarriors|url=http://www.marketplace.org/topics/tech/human-brain-project-and-recruiting-more-cyberwarriors|access-date=22 February 2013|date=January 29, 2013}}</ref><br />
<br />
'''<font color="#ff8000">人类大脑计划Human Brain Project</font>'''对神经形态工程具有较大影响，其主要任务是尝试用生物数据在超级计算机中模拟完整的人脑。人类大脑计划由神经科学、医学和计算机科学背景的研究人员组成。<ref name=":18" />该项目的联合主管亨利•马克拉姆(Henry Markram)表示，人类大脑计划的目的是建立一个探索和了解脑科学和脑疾病知识的基础，并利用这些知识来构建更先进的计算机技术。这个项目的三个主要目标分别是: 更好地理解大脑的各个部分是如何相互配合协同工作的; 理解如何客观地诊断和治疗脑部疾病; 以及利用对人类大脑的理解来开发神经形态计算机。模拟一个完整的人类大脑需要一台比现在强大一千倍的超级计算机，这不断激发着对神经形态计算机领域的研究兴趣。<ref name=":19" />欧盟委员会已经向人类大脑计划拨款13亿美元。<ref name=":20" /><br />
<br />
Other research with implications for neuromorphic engineering involves the [[BRAIN Initiative]]<ref name="economist">[https://www.economist.com/news/science-and-technology/21582495-computers-will-help-people-understand-brains-better-and-understanding-brains Neuromorphic computing: The machine of a new soul], The Economist, 2013-08-03</ref> and the [[TrueNorth]] chip from [[IBM]].<ref name=":21">{{cite journal|last1=Modha|first1=Dharmendra|title=A million spiking-neuron integrated circuit with a scalable communication network and interface|journal=Science|date=Aug 2014|volume=345|issue=6197|pages=668–673|doi=10.1126/science.1254642|pmid=25104385|bibcode=2014Sci...345..668M|s2cid=12706847}}</ref> Neuromorphic devices have also been demonstrated using nanocrystals, nanowires, and conducting polymers.<ref name=":22">{{Cite web|url=http://jessamynfairfield.com/wp-content/uploads/2017/03/PWMar17Fairfield.pdf|title=Smarter Machines|last=Fairfield|first=Jessamyn|date=March 1, 2017}}</ref><br />
<br />
其他与神经形态工程有关的研究还包括'''<font color="#ff8000">脑计划BRAIN initiative</font>'''，<ref name="economist" />和IBM研发的'''<font color="#ff8000">TrueNorth</font>'''芯片。<ref name=":21" />使用纳米晶体、纳米线和导电聚合物也能够用于制造神经形态学硬件。<ref name=":22" /><br />
<br />
[[Intel]] unveiled its neuromorphic research chip, called “[[Intel Loihi|Loihi]]”, in October 2017. The chip uses an asynchronous [[spiking neural network]] (SNN) to implement adaptive self-modifying event-driven fine-grained parallel computations used to implement learning and inference with high efficiency.<ref name=":23">{{cite journal |last1=Davies |first1=Mike |title=Loihi: A Neuromorphic Manycore Processor with On-Chip Learning |journal=IEEE Micro |date=January 16, 2018 |volume=38 |issue=1 |pages=82–99 |display-authors=etal|doi=10.1109/MM.2018.112130359 |s2cid=3608458 }}</ref><ref name="Morris2017">{{cite web |last1=Morris |first1=John |title=Why Intel built a neuromorphic chip |url=https://www.zdnet.com/article/why-intel-built-a-neuromorphic-chip/ |website=ZDNet |access-date=17 August 2018 |language=en}}</ref><br />
<br />
2017年10月，英特尔发布了神经形态芯片'''<font color="#ff8000">Loihi</font>'''。该芯片采用异步'''<font color="#ff8000">脉冲神经网络Spiking neural network</font>'''实现了自适应、自修改、事件驱动的细粒度并行计算，实现了高效的学习和推理。<ref name=":23" /><ref name="Morris2017" /><br />
<br />
[[IMEC]], a Belgium-based nanoelectronics research center, demonstrated the world's first self-learning neuromorphic chip. The brain-inspired chip, based on OxRAM technology, has the capability of self-learning and has been demonstrated to have the ability to compose music.<ref name=":24">{{cite web |title=Imec demonstrates self-learning neuromorphic chip that composes music |url=https://www.imec-int.com/en/articles/imec-demonstrates-self-learning-neuromorphic-chip-that-composes-music |website=IMEC International |access-date=1 October 2019}}</ref> IMEC released the 30-second tune composed by the prototype. The chip was sequentially loaded with songs in the same time signature and style. The songs were old Belgian and French flute minuets, from which the chip learned the rules at play and then applied them.<ref name=":25">{{cite web|last1=Bourzac|first1=Katherine|title=A Neuromorphic Chip That Makes Music|url=https://spectrum.ieee.org/a-neuromorphic-chip-that-makes-music|url-status=live|access-date=1 October 2019|website=IEEE Spectrum|date=May 23, 2017}}</ref><br />
<br />
比利时的'''<font color="#ff8000">微电子研究中心IMEC</font>'''研发了世界上首个自学习神经形态芯片。这种基于'''<font color="#32CD32"> OxRAM(filamentary-oxide-based resistive memory technology)</font>''' 技术的大脑启发芯片具有自学习能力，并且已被证明具有创作音乐的能力。<ref name=":24" />IMEC发布了由芯片原型机谱写的30秒旋律。向芯片加载一系列特征、风格统一的歌曲（古代比利时和法国长笛小步舞曲），芯片从中学习相关规则并将其应用于创作。<ref name=":25" /><br />
<br />
[[Blue Brain Project|The Blue Brain Project]], led by Henry Markram, aims to build biologically detailed digital reconstructions and simulations of the mouse brain. The Blue Brain Project has created in silico models of rodent brains, while attempting to replicate as many details about its biology as possible. The supercomputer-based simulations offer new perspectives on understanding the structure and functions of the brain.<br />
<br />
由亨利·马克拉姆领导的'''<font color="#ff8000">蓝脑计划The Blue Brain Project</font>'''旨在建立小鼠大脑生理学细节的数字重建和模拟。蓝脑计划已经建立了啮齿动物大脑的电子模型，同时进行着尽可能多地复制其生理学细节的尝试。基于超级计算机的模拟为理解大脑的结构和功能提供了新的视角。<br />
<br />
The European Union funded a series of projects at the University of Heidelberg, which led to the development of <br />
[[BrainScaleS]] (brain-inspired multiscale computation in neuromorphic hybrid systems), a hybrid analog [[neuromorphic]] supercomputer located at Heidelberg University, Germany. It was developed as part of the [[Human Brain Project]] neuromorphic computing platform and is the complement to the [[SpiNNaker]] supercomputer (which is based on digital technology). The architecture used in BrainScaleS mimics biological neurons and their connections on a physical level; additionally, since the components are made of silicon, these model neurons operate on average 864 times (24 hours of real time is 100 seconds in the machine simulation) that of their biological counterparts.<ref name=":26">{{Cite web|date=2016-03-21|title=Beyond von Neumann, Neuromorphic Computing Steadily Advances|url=https://www.hpcwire.com/2016/03/21/lacking-breakthrough-neuromorphic-computing-steadily-advance/|access-date=2021-10-08|website=HPCwire|language=en-US}}</ref><br />
<br />
欧盟资助了海德堡大学一系列促进BrainScaleS(神经形态混合系统中受大脑启发的多尺度计算)发展的项目，这是一台位于德国海德堡大学的混合模拟'''<font color="#ff8000">神经形态Neuromorphic</font>'''超级计算机。它是作为人类大脑计划中神经形态计算平台的一部分而开发的，是'''<font color="#ff8000">SpiNNaker</font>'''超级计算机(基于数字技术)的补充。BrainScaleS中使用的体系架构模拟了生物神经元及其在物理层面上的连接；此外，由于这些组件是由硅制成的，这些模型神经元平均运行速度是生物神经元的864倍，这意味着在机器模拟中，24小时的实时时间仅为100秒。<ref name=":26" /><br />
<br />
===Neuromorphic sensors===<br />
===神经形态传感器===<br />
The concept of neuromorphic systems can be extended to sensors (not just to computation). An example of this applied to detecting [[light]] is the [[retinomorphic sensor]] or, when employed in an array, the [[event camera]].<br />
<br />
神经形态系统的概念可以扩展到传感器(而不仅仅是计算单元)。一个用于检测光线的例子是'''<font color="#ff8000">类视网膜传感器Retinomorphic sensor</font>'''，或者'''<font color="#ff8000">事件摄像机Event camera</font>阵列。'''<br />
<br />
<br />
==Ethical considerations ==<br />
==伦理问题==<br />
While the interdisciplinary concept of neuromorphic engineering is relatively new, many of the same ethical considerations apply to neuromorphic systems as apply to [[human-like machines]] and [[artificial intelligence]] in general. However, the fact that neuromorphic systems are designed to mimic a [[human brain]] gives rise to unique ethical questions surrounding their usage.<br />
<br />
虽然神经形态工程这个跨学科概念相对较新，但许多适用于'''<font color="#ff8000">类人机器Human-like machines</font>'''和人工智能的伦理考虑大体上也适用于神经形态系统。另一方面，神经形态系统是为了模仿人类大脑而设计，这一事实引起了一些针对性的独特伦理问题。<br />
<br />
However, the practical debate is that neuromorphic hardware as well as artificial "neural networks" are immensely simplified models of how the brain operates or processes information at a much lower [[complex system|complexity in terms of size and functional technology]] and a much more regular structure in terms of [[brain connectivity|connectivity]]. Comparing [[neuromorphic chip]]s to the brain is a very crude comparison similar to comparing a plane to a bird just because they both have wings and a tail. The fact is that neural cognitive systems are many orders of magnitude more [[Energy efficiency (physics)|energy-]] and compute-efficient than current state-of-the-art AI and neuromorphic engineering is an attempt to narrow this gap by inspiring from the brain's mechanism just like many engineering designs have [[bioengineering|bio-inspired features]].<br />
<br />
然而，这种争论的实际情况是，神经形态硬件和人工“神经网络”是对大脑运作或处理信息过程的极其简化的模型，在大小和功能技术方面的'''<font color="#ff8000">复杂性Complexity</font>'''相比而言比较低，在'''<font color="#ff8000">连接性Connectivity</font>'''方面也具有更加规则的结构。将'''<font color="#ff8000">神经形态芯片Neuromorphic chips</font>'''与大脑进行比较是一种非常粗糙的比较，类似于仅仅因为飞机有翅膀和尾巴就将它与鸟进行比较。事实上，当前最先进的人工智能在'''<font color="#ff8000">能耗效率Energy-efficiency</font>'''和计算效率方面距离人脑神经认知系统仍有较大差距，而神经形态工程只是一种通过从大脑机制中获得灵感来缩小这种差距的尝试，就像许多工程设计中都具有'''<font color="#ff8000">生物启发的特征Bio-inspired features</font>'''一样。<br />
<br />
===Democratic concerns===<br />
===公众担忧===<br />
Significant ethical limitations may be placed on neuromorphic engineering due to public perception.<ref name=":27">{{Cite report|url=https://ai100.stanford.edu/sites/g/files/sbiybj9861/f/ai_100_report_0831fnl.pdf|title=Artificial Intelligence and Life in 2030|author=2015 Study Panel|date=September 2016|work=One Hundred Year Study on Artificial Intelligence (AI100)|publisher=Stanford University}}</ref> Special [[Eurobarometer]] 382: Public Attitudes Towards Robots, a survey conducted by the European Commission, found that 60% of [[European Union]] citizens wanted a ban of robots in the care of children, the elderly, or the disabled. Furthermore, 34% were in favor of a ban on robots in education, 27% in healthcare, and 20% in leisure. The European Commission classifies these areas as notably “human.” The report cites increased public concern with robots that are able to mimic or replicate human functions. Neuromorphic engineering, by definition, is designed to replicate the function of the human brain.<ref name=":1">{{Cite web|url=http://ec.europa.eu/commfrontoffice/publicopinion/archives/ebs/ebs_382_en.pdf|title=Special Eurobarometer 382: Public Attitudes Towards Robots|last=European Commission|date=September 2012|website=European Commission}}</ref><br />
<br />
由于公众认知的相关忧虑，神经形态工程学可能会受到严重的伦理限制。<ref name=":27" />欧盟委员会进行的一项调查发现，60% 的欧盟公民希望禁止机器人参与照顾儿童、老人或残疾人的工作。此外，34% 的人支持禁止机器人用于教育，27% 的人支持禁止机器人用于医疗保健，20% 的人支持禁止机器人用于休闲。欧盟委员会将以上领域划入“人类”范畴。报告指出，公众越来越关注能够模仿或复制人类功能的机器人。神经形态工程，顾名思义，是为了复制人脑的功能而设计的。<ref name=":1" /><br />
<br />
The democratic concerns surrounding neuromorphic engineering are likely to become even more profound in the future. The European Commission found that EU citizens between the ages of 15 and 24 are more likely to think of robots as human-like (as opposed to instrument-like) than EU citizens over the age of 55. When presented an image of a robot that had been defined as human-like, 75% of EU citizens aged 15–24 said it corresponded with the idea they had of robots while only 57% of EU citizens over the age of 55 responded the same way. The human-like nature of neuromorphic systems, therefore, could place them in the categories of robots many EU citizens would like to see banned in the future.<ref name=":1" /><br />
<br />
围绕神经形态工程的公众担忧可能在未来变得更加深刻。欧盟委员会发现，相比于55岁以上的欧盟公民，15至24岁的欧盟公民更有可能认为机器人像人(而不是像仪器)。当看到一张“类人”机器人的图片时，年龄在15岁至24岁之间的欧盟公民中有75% 的人表示这符合他们对机器人的想法，而55岁以上的欧盟公民中只有57% 的人有同样的反应。因此，神经形态系统可能因为其类似人类的特性而被归入许多欧盟公民希望在未来禁止使用的机器人类别。<ref name=":1" /><br />
<br />
=== Personhood===<br />
=== 人格问题===<br />
As neuromorphic systems have become increasingly advanced, some scholars{{who|date=August 2021}} have advocated for granting [[personhood]] rights to these systems. If the brain is what grants humans their personhood, to what extent does a neuromorphic system have to mimic the human brain to be granted personhood rights? Critics of technology development in the [[Human Brain Project]], which aims to advance brain-inspired computing, have argued that advancement in neuromorphic computing could lead to machine consciousness or personhood.<ref name=":28">{{Cite journal|last=Aicardi|first=Christine|date=September 2018|title=Accompanying technology development in the Human Brain Project: From foresight to ethics management|journal=Futures|volume=102|pages=114–124|doi=10.1016/j.futures.2018.01.005|doi-access=free}}</ref> If these systems are to be treated as people, critics argue, then many tasks humans perform using neuromorphic systems, including the act of termination of neuromorphic systems, may be morally impermissible as these acts would violate the autonomy of the neuromorphic systems.<ref name=":29">{{Cite journal|last=Lim|first=Daniel|date=2014-06-01|title=Brain simulation and personhood: a concern with the Human Brain Project|journal=Ethics and Information Technology|language=en|volume=16|issue=2|pages=77–89|doi=10.1007/s10676-013-9330-5|s2cid=17415814|issn=1572-8439}}</ref><br />
<br />
随着神经形态系统的日益发展，一些学者主张赋予这些系统'''<font color="ff8000">人格Personhood</font>'''权。如果是大脑赋予了人类人格，那么在多大程度上模仿人类大脑的神经形态系统才能被赋予人格权利？“人类大脑计划”旨在推进以大脑为灵感的计算机技术发展，该计划的批评者认为，神经形态计算机技术的进步可能导致机器意识或人格的形成。<ref name=":28" />这些批评者认为，如果这些系统被当作人来对待，那么人类使用神经形态系统执行任务（包括终止神经形态系统）的行为，在道德上就可能是不被允许的，因为这些行为将违反神经形态系统的自主性。<ref name=":29" /><br />
<br />
<br />
==Dual use (military applications)==<br />
==军民两用技术（军事应用）==<br />
The [[Joint Artificial Intelligence Center]], a branch of the U.S. military, is a center dedicated to the procurement and implementation of AI software and neuromorphic hardware for combat use. Specific applications include smart headsets/goggles and robots. JAIC intends to rely heavily on neuromorphic technology to connect "every fighter every shooter" within a network of neuromorphic-enabled units.<br />
<br />
'''<font color="#ff8000">联合人工智能中心The Joint Artificial Intelligence Center</font>'''(JAIC),是美国军队的一个分支，专门从事采购和实施用于战斗的人工智能软件和神经形态硬件。具体应用包括智能耳机/护目镜和机器人。JAIC打算高度依赖神经形态技术，'''<font color="#32CD32">使用神经形态技术来连接神经形态单位网络中的“每个战士每个射手”</font>'''。<br />
<br />
<br />
==Legal considerations==<br />
==法律问题==<br />
Skeptics have argued that there is no way to apply the electronic personhood, the concept of personhood that would apply to neuromorphic technology, legally. In a letter signed by 285 experts in law, robotics, medicine, and ethics opposing a European Commission proposal to recognize “smart robots” as legal persons, the authors write, “A legal status for a robot can’t derive from the [[Natural person|Natural Person]] model, since the robot would then hold [[human rights]], such as the right to dignity, the right to its integrity, the right to remuneration or the right to citizenship, thus directly confronting the Human rights. This would be in contradiction with the [[Charter of Fundamental Rights of the European Union]] and the [[Convention for the Protection of Human Rights and Fundamental Freedoms]].”<ref name=":30">{{Cite web|url=http://www.robotics-openletter.eu/|title=Robotics Openletter {{!}} Open letter to the European Commission|language=fr-FR|access-date=2019-05-10}}</ref><br />
<br />
质疑者者认为，在法律上没有办法应用能够适用于神经形态技术的电子人格。在一封由285名法律、机器人技术、医学和伦理学专家的联名信中，作者们反对欧盟委员会提出的承认“智能机器人”为法人的提议。他们写道，“机器人的法律地位无法从'''<font color="#ff8000">自然人Natural Person</font>'''模型中推导出来，因为机器人将被赋予'''<font color="#ff8000">人权Human rights</font>'''，如尊严权、完整权、报酬权或公民权，从而直接面临人权问题。这将有悖于《欧联基本权利宪章》和《欧洲保障人权和根本自由公约》”。<ref name=":30" /><br />
<br />
===Ownership and property rights ===<br />
===所有权及财产权问题===<br />
There is significant legal debate around property rights and artificial intelligence. In ''Acohs Pty Ltd v. Ucorp Pty Ltd'', Justice Christopher Jessup of the [[Federal Court of Australia]] found that the [[source code]] for [[Material safety data sheets|Material Safety Data Sheets]] could not be [[Copyright law of Australia|copyrighted]] as it was generated by a [[software interface]] rather than a human author.<ref name=":31">{{Cite web|url=http://www.lavan.com.au/advice/intellectual_property_technology/copyright_in_source_code_and_digital_products|title=Copyright in source code and digital products|last=Lavan|website=Lavan|language=en|access-date=2019-05-10}}</ref> The same question may apply to neuromorphic systems: if a neuromorphic system successfully mimics a human brain and produces a piece of original work, who, if anyone, should be able to claim ownership of the work?<ref name=":32">{{cite journal |last1=Eshraghian|first1=Jason K. |title=Human Ownership of Artificial Creativity |journal=Nature Machine Intelligence |date=9 March 2020 |volume=2 |pages=157–160 |doi=10.1038/s42256-020-0161-x}}</ref><br />
<br />
法律界围绕财产权和人工智能有着重大争论。在Acohs Pty有限公司诉Ucorp Pty有限公司一案中，澳大利亚联邦法院的克里斯托弗·杰瑟普法官发现，版权保护不适用于材料安全数据表的'''<font color="#ff8000">源代码Source code</font>'''，因为它是由'''<font color="#ff8000">软件界面Software interface</font>'''生成而非人类工作者生成的。<ref name=":31" />同样的问题可能也适用于神经形态系统: 如果一个神经形态系统成功地模仿了人类的大脑，并产生了一部原创作品，那么该如何确认这部作品的所有权归属？<ref name=":32" /><br />
<br />
<br />
==Neuromemristive systems==<br />
==神经记忆电阻系统==<br />
Neuromemristive systems are a subclass of neuromorphic computing systems that focus on the use of [[memristors]] to implement [[neuroplasticity]]. While neuromorphic engineering focuses on mimicking biological behavior, neuromemristive systems focus on abstraction.<ref name=":33">{{Cite web|url=https://digitalops.sandia.gov/Mediasite/Play/a10cf6ceb55d47608bb8326dd00e46611d|title=002.08 N.I.C.E. Workshop 2014: Towards Intelligent Computing with Neuromemristive Circuits and Systems - Feb. 2014|website=digitalops.sandia.gov|access-date=2019-08-26}}</ref> For example, a neuromemristive system may replace the details of a [[Cerebral cortex|cortical]] microcircuit's behavior with an abstract neural network model.<ref name=":34">C. Merkel and D. Kudithipudi, "Neuromemristive extreme learning machines for pattern classification," ISVLSI, 2014.</ref><br />
<br />
神经记忆电阻系统是神经形态计算系统的一个亚类，主要研究利用'''<font color="ff8000">记忆电阻器Memristors</font>'''实现'''<font color="ff8000">神经可塑性Neuroplasticity</font>'''。神经形态工程的重点是模拟生物行为，而神经记忆电阻系统的重点是提取。<ref name=":33" />举个例子，一个神经记忆系统可能用抽象的神经网络模型替代'''<font color="ff8000">皮层Cortical</font>'''微电路的行为细节。<ref name=":34" /><br />
<br />
There exist several neuron inspired threshold logic functions<ref name="Maan 1–13" /> implemented with memristors that have applications in high level [[pattern recognition]] applications. Some of the applications reported recently include [[speech recognition]],<ref name=":35">{{Cite journal|title = Memristor pattern recogniser: isolated speech word recognition|journal = Electronics Letters|pages = 1370–1372|volume = 51|issue = 17|doi = 10.1049/el.2015.1428|first1 = A.K.|last1 = Maan|first2 = A.P.|last2 = James|first3 = S.|last3 = Dimitrijev|year = 2015|bibcode = 2015ElL....51.1370M|hdl = 10072/140989|s2cid = 61454815|url = https://semanticscholar.org/paper/48d3ab11ec6e213b62f11eedcfb7b7febb058674|hdl-access = free}}</ref> [[face recognition]]<ref name=":36">{{Cite journal|title = Memristive Threshold Logic Face Recognition|journal = Procedia Computer Science|date = 2014-01-01|pages = 98–103|volume = 41|series = 5th Annual International Conference on Biologically Inspired Cognitive Architectures, 2014 BICA|doi = 10.1016/j.procs.2014.11.090|first1 = Akshay Kumar|last1 = Maan|first2 = Dinesh S.|last2 = Kumar|first3 = Alex Pappachen|last3 = James|doi-access = free}}</ref> and [[object recognition]].<ref name=":37">{{Cite journal|title = Memristive Threshold Logic Circuit Design of Fast Moving Object Detection|journal = IEEE Transactions on Very Large Scale Integration (VLSI) Systems|date = 2015-10-01|issn = 1063-8210|pages = 2337–2341|volume = 23|issue = 10|doi = 10.1109/TVLSI.2014.2359801|first1 = A.K.|last1 = Maan|first2 = D.S.|last2 = Kumar|first3 = S.|last3 = Sugathan|first4 = A.P.|last4 = James|arxiv = 1410.1267|s2cid = 9647290}}</ref> They also find applications in replacing conventional digital logic gates.<ref name=":38">{{Cite journal|title = Resistive Threshold Logic|journal = IEEE Transactions on Very Large Scale Integration (VLSI) Systems|date = 2014-01-01|issn = 1063-8210|pages = 190–195|volume = 22|issue = 1|doi = 10.1109/TVLSI.2012.2232946|first1 = A.P.|last1 = James|first2 = L.R.V.J.|last2 = Francis|first3 = D.S.|last3 = Kumar|arxiv = 1308.0090|s2cid = 7357110}}</ref><ref name=":39">{{Cite journal|title = Threshold Logic Computing: Memristive-CMOS Circuits for Fast Fourier Transform and Vedic Multiplication|journal = IEEE Transactions on Very Large Scale Integration (VLSI) Systems|date = 2015-11-01|issn = 1063-8210|pages = 2690–2694|volume = 23|issue = 11|doi = 10.1109/TVLSI.2014.2371857|first1 = A.P.|last1 = James|first2 = D.S.|last2 = Kumar|first3 = A.|last3 = Ajayan|arxiv = 1411.5255|s2cid = 6076956}}</ref><br />
<br />
受神经元启发、使用记忆电阻器实现的阈值逻辑函数<ref name="Maan 1–13" />在高级'''<font color="#ff8000">模式识别Pattern recognition</font>'''中有着广泛的应用，最近报道中其应用包括'''<font color="#ff8000">语音识别Speech recognition<ref name=":35" /></font>'''、'''<font color="#ff8000">人脸识别Face recognition<ref name=":36" /></font>'''和'''<font color="#ff8000">物体识别Object recognition<ref name=":37" /></font>'''。阈值逻辑函数还可以用来取代传统的数字逻辑门。<ref name=":38" /><ref name=":39" /><br />
<br />
For ideal passive memristive circuits there is an exact equation (Caravelli-Traversa-[[Di Ventra]] equation) for the internal memory of the circuit:<ref name=":40">{{cite journal |last=Caravelli |display-authors=etal|arxiv=1608.08651 |title=The complex dynamics of memristive circuits: analytical results and universal slow relaxation |year=2017 |doi=10.1103/PhysRevE.95.022140 |pmid= 28297937 |volume=95 |issue= 2 |pages= 022140 |journal=Physical Review E|bibcode=2017PhRvE..95b2140C |s2cid=6758362}}</ref><br />
<br />
对于理想的无源记忆电路，电路的内部记忆可以用精确的方程(Caravelli-Traversa-Di Ventra方程) 来描述:<ref name=":40" /><br />
<br />
:<math> \frac{d}{dt} \vec{W} = \alpha \vec{W}-\frac{1}{\beta} (I+\xi \Omega W)^{-1} \Omega \vec S </math><br />
<br />
as a function of the properties of the physical memristive network and the external sources. In the equation above, <math>\alpha</math> is the "forgetting" time scale constant, <math> \xi=r-1</math> and <math>r=\frac{R_\text{off}}{R_\text{on}}</math> is the ratio of ''off'' and ''on'' values of the limit resistances of the memristors, <math> \vec S </math> is the vector of the sources of the circuit and <math>\Omega</math> is a projector on the fundamental loops of the circuit. The constant <math>\beta</math> has the dimension of a voltage and is associated to the properties of the [[memristor]]; its physical origin is the charge mobility in the conductor. The diagonal matrix and vector <math>W=\operatorname{diag}(\vec W)</math> and <math>\vec W<\math> respectively, are instead the internal value of the memristors, with values between 0 and 1. This equation thus requires adding extra constraints on the memory values in order to be reliable.<br />
<br />
Caravelli-Traversa-Di Ventra方程是描述物理记忆网络和外部源性质的函数。在上述方程中，<math>\alpha</math>是“遗忘”时间尺度常数，<math>\xi=r-1</math>，<math>r =\frac{R\text_{off}}{R_\text{on}}</math>是记忆电阻器off状态和on状态极限电阻值之比，<math>\vec S</math>是电路源的矢量，<math>\Omega</math>是电路基本环路的投影。常数<math>\beta</math>具有电压的量纲，与记忆电阻器的特性有关；其物理原型是导体中的电荷迁移率。对角矩阵和向量 <math>W=\operatorname{diag}(\vec W)</math>和<math>\vec W</math>'''<font color="#32CD32">分别是记忆电阻器的内阻</font>'''，值在0到1之间。因此，这个等式需要在'''<font color="32CD32">内存值</font>'''上添加额外约束以保证可靠性。<br />
<br />
<br />
==See also==<br />
==相关词条==<br />
{{Columns-list|colwidth=18em|<br />
* [[AI accelerator (computer hardware)|AI accelerator]]<br />
* [[Artificial brain]]<br />
* [[Biomorphic]]<br />
* [[Cognitive computer]]<br />
* [[Computation and Neural Systems]]<br />
* [[Differentiable programming]]<br />
* [[Event camera]]<br />
* [[Neurorobotics]]<br />
* [[Optical flow sensor]]<br />
* [[Physical neural network]]<br />
* [[SpiNNaker]]<br />
* [[SyNAPSE]]<br />
* [[Retinomorphic sensor]]<br />
* [[Vision chip]]<br />
* [[Vision processing unit]]<br />
* [[Zeroth (software)]]<br />
* [[Hardware for artificial intelligence]]<br />
}}<br />
<br />
{{Portal bar|Electronics}}<br />
<br />
==References==<br />
{{Reflist|40em}}<br />
<br />
==External links==<br />
<!--======================== {{No more links}} ============================<br />
| PLEASE BE CAUTIOUS IN ADDING MORE LINKS TO THIS ARTICLE. Wikipedia |<br />
| is not a collection of links nor should it be used for advertising. |<br />
| |<br />
| Excessive or inappropriate links WILL BE DELETED. |<br />
| See [[Wikipedia:External links]] & [[Wikipedia:Spam]] for details. |<br />
| |<br />
| If there are already plentiful links, please propose additions or |<br />
| replacements on this article's discussion page, or submit your link |<br />
| to the relevant category at the Open Directory Project (dmoz.org) |<br />
| and link back to that category using the {{dmoz}} template. |<br />
======================={{No more links}}=============================--><br />
{{Commons category|Neuromorphic Engineering}}<br />
*[https://web.archive.org/web/20150727034331/http://ine-web.org/workshops/workshops-overview Telluride Neuromorphic Engineering Workshop]<br />
*[https://archive.today/20130115190057/http://capocaccia.ethz.ch/ CapoCaccia Cognitive Neuromorphic Engineering Workshop]<br />
*[https://web.archive.org/web/20190716132350/http://www.ine-web.org/ Institute of Neuromorphic Engineering]<br />
*[http://www.ine-news.org/ INE news site].<br />
*[http://www.frontiersin.org/neuromorphic_engineering Frontiers in Neuromorphic Engineering Journal]<br />
*[http://www.cns.caltech.edu/ Computation and Neural Systems] department at the [[California Institute of Technology]].<br />
*[http://www.humanbrainproject.eu/ Human Brain Project official site]<br />
*[https://www.the-scientist.com/features/building-a-silicon-brain-65738 Building a Silicon Brain:] Computer chips based on biological neurons may help simulate larger and more-complex brain models. May 1, 2019. SANDEEP RAVINDRAN<br />
<br />
<br />
*Telluride Neuromorphic Engineering Workshop<br />
* CapoCaccia Cognitive Neuromorphic Engineering Workshop<br />
* Institute of Neuromorphic Engineering<br />
*INE news site.<br />
*Frontiers in Neuromorphic Engineering Journal<br />
*Computation and Neural Systems department at the California Institute of Technology.<br />
*Human Brain Project official site<br />
*Building a Silicon Brain: Computer chips based on biological neurons may help simulate larger and more-complex brain models. May 1, 2019. SANDEEP RAVINDRAN<br />
<br />
=<nowiki>外部链接</nowiki>= <br />
*碲化物神经形态工程工作室<br />
*CapoCaccia认知神经形态工程工作室<br />
*神经形态工程研究所<br />
*INE新闻站点。<br />
*《神经形态工程学前沿》<br />
*加州理工学院计算与神经系统系。<br />
*人脑项目官方网站<br />
*打造硅脑: 基于生物神经元的计算机芯片可能有助于模拟更大、更复杂的大脑模型。2019年5月1日。SANDEEP RAVINDRAN.<br />
<br />
{{Differentiable computing}}<br />
{{Authority control}}<br />
<br />
[[Category:Electrical engineering]]<br />
[[Category:Neuroscience]]<br />
[[分类:AI accelerators|*]]<br />
[[Category:Artificial intelligence]]<br />
[[Category:Robotics]]<br />
<br />
Category:Electrical engineering<br />
Category:Neuroscience<br />
*<br />
Category:Artificial intelligence<br />
Category:Robotics<br />
<br />
类别: 电气工程类别: 神经科学 * 类别: 人工智能类别: 机器人<br />
<br />
<noinclude><br />
<br />
<small>This page was moved from [[wikipedia:en:Neuromorphic engineering]]. Its edit history can be viewed at [[神经形态计算/edithistory]]</small></noinclude><br />
<br />
[[Category:待整理页面]]<br />
<br />
<references /></div>Shenky20https://wiki.swarma.org/index.php?title=%E7%A5%9E%E7%BB%8F%E5%BD%A2%E6%80%81%E8%AE%A1%E7%AE%97&diff=30751神经形态计算2022-04-28T11:50:59Z<p>Shenky20：/* 神经记忆电阻系统 */</p>
<hr />
<div><br />
此词条由神经动力学模型读书会词条梳理志愿者Shenky20翻译审校，未经专家审校，带来阅读不便，请见谅。<br />
<br />
{{Use American English|date = January 2019}}<br />
{{Use mdy dates|date = January 2019}}<br />
<br />
'''Neuromorphic engineering''', also known as '''neuromorphic computing''',<ref name=":3">{{Cite journal | doi = 10.1145/2601069| title = Neuromorphic computing gets ready for the (really) big time| journal = [[Communications of the ACM]]| volume = 57| issue = 6| year = 2014| last1 = Monroe | first1 = D. | pages = 13–15| s2cid = 20051102}}</ref><ref name=":4">{{Cite journal | doi = 10.1088/0957-4484/21/17/175202| pmid = 20368686| title = Nanotube devices based crossbar architecture: Toward neuromorphic computing| journal = Nanotechnology| volume = 21| issue = 17| pages = 175202| year = 2010| last1 = Zhao | first1 = W. S. | last2 = Agnus | first2 = G. | last3 = Derycke | first3 = V. | last4 = Filoramo | first4 = A. | last5 = Bourgoin | first5 = J. -P. | last6 = Gamrat | first6 = C. | bibcode = 2010Nanot..21q5202Z| url = https://zenodo.org/record/3428659}}</ref><ref name="humanbrainproject">{{YouTube|id=6RoiZ90mGfw|title=The Human Brain Project SP 9: Neuromorphic Computing Platform}}</ref> is the use of [[very-large-scale integration]] (VLSI) systems containing electronic [[analog circuit]]s to mimic neuro-biological architectures present in the nervous system. A neuromorphic computer/chip is any device that uses physical artificial neurons (made from silicon) to do computations.<ref name=":5">{{cite journal|last1=Mead|first1=Carver|title=Neuromorphic electronic systems|journal=Proceedings of the IEEE|date=1990|volume=78|issue=10|pages=1629–1636|doi=10.1109/5.58356|url=https://authors.library.caltech.edu/53090/1/00058356.pdf}}</ref><ref name=":2" /> In recent times, the term ''neuromorphic'' has been used to describe analog, digital, mixed-mode analog/digital VLSI, and software systems that implement models of [[neural system]]s (for [[perception]], [[motor control]], or [[multisensory integration]]). The implementation of neuromorphic computing on the hardware level can be realized by oxide-based [[memristor]]s,<ref name="Maan 1–13">{{Cite journal|last1=Maan|first1=A. K.|last2=Jayadevi|first2=D. A.|last3=James|first3=A. P.|date=2016-01-01|title=A Survey of Memristive Threshold Logic Circuits|journal=IEEE Transactions on Neural Networks and Learning Systems|volume=PP|issue=99|pages=1734–1746|doi=10.1109/TNNLS.2016.2547842|pmid=27164608|issn=2162-237X|arxiv=1604.07121|bibcode=2016arXiv160407121M|s2cid=1798273}}</ref> spintronic memories, threshold switches, and [[transistor]]s.<ref name=":6">{{Cite journal|title = Mott Memory and Neuromorphic Devices|journal = Proceedings of the IEEE|date = 2015-08-01|issn = 0018-9219|pages = 1289–1310|volume = 103|issue = 8|doi = 10.1109/JPROC.2015.2431914|first1 = You|last1 = Zhou|first2 = S.|last2 = Ramanathan|s2cid = 11347598|url = https://zenodo.org/record/895565}}</ref><ref name=":2">{{Cite document|title=Neuromorphic Circuits With Neural Modulation Enhancing the Information Content of Neural Signaling {{!}} International Conference on Neuromorphic Systems 2020|language=EN|doi=10.1145/3407197.3407204|s2cid=220794387}}</ref> Training software-based neuromorphic systems of spiking neural networks can be achieved using error backpropagation, e.g., using [[Python (programming language)|Python]] based frameworks such as snnTorch,<ref name=":7">{{cite journal |last1=Eshraghian|first1=Jason K.|last2=Ward|first2=Max|last3=Neftci |first3=Emre|last4=Wang|first4=Xinxin|last5=Lenz|first5=Gregor|last6=Dwivedi|first6=Girish|last7=Bennamoun|first7=Mohammed|last8=Jeong|first8=Doo Seok|last9=Lu|first9=Wei D.|title=Training Spiking Neural Networks Using Lessons from Deep Learning |date=1 October 2021 |arxiv=2109.12894 }}</ref> or using canonical learning rules from the biological learning literature, e.g., using BindsNet.<ref name=":8">{{Cite web | url=https://github.com/Hananel-Hazan/bindsnet | title=Hananel-Hazan/bindsnet: Simulation of spiking neural networks (SNNs) using PyTorch.| date=31 March 2020}}</ref><br />
<br />
<br />
'''<font color="#ff8000">神经形态工程Neuromorphic engineering</font>'''(也称为'''<font color="#ff8000">神经形态计算Neuromorphic computing</font>''')<ref name=":3" /><ref name=":4" /><ref name="humanbrainproject" />是指使用包含电子'''<font color="#ff8000">模拟电路Analog circuit</font>'''的'''<font color="#ff8000">超大规模集成电路Very-large-scale integration</font>'''系统来模拟神经系统中生理结构的研究方法。神经形态计算机或神经形态芯片包括任何使用由硅制成的人造神经元进行计算的设备。<ref name=":5" /><ref name=":2" />近年来，神经形态学（neuromorphic）这个术语被用来描述能够实现'''<font color="#ff8000">神经系统Neural system</font>'''模型功能(如'''<font color="#ff8000">感知Perception</font>'''、'''<font color="#ff8000">运动控制Motor control</font>'''，'''<font color="#ff8000">多感官整合Multisensory integration</font>'''等)的模拟、数字、模拟/数字混合模式超大规模集成电路和软件系统。神经形态计算的硬件实现可以通过基于氧化物的'''<font color="#ff8000">记忆电阻器Memristor</font>'''、<ref name="Maan 1–13" />自旋电子存储器、阈值开关和'''<font color="#ff8000">晶体管Transistor</font>'''来实现。<ref name=":6" /><ref name=":2" />对基于软件的脉冲神经网络系统的训练可以通过误差反向传播机制来实现，例如，使用snnTorch等基于Python的框架，<ref name=":7" />或使用BindsNet等典型的受生物启发的学习模式。<ref name=":8" /><br />
<br />
A key aspect of neuromorphic engineering is understanding how the morphology of individual neurons, circuits, applications, and overall architectures creates desirable computations, affects how information is represented, influences robustness to damage, incorporates learning and development, adapts to local change (plasticity), and facilitates evolutionary change.<br />
<br />
<br />
神经形态工程领域的一个关键问题，就是理解'''<font color="#32CD32">单个神经元形态、神经回路、应用和整体结构</font>'''如何产生理想的计算，如何影响信息的表示和对破坏的鲁棒性，如何整合学习和发展，如何适应局部变化(可塑性) 并促进逐渐发展的变化。<br />
<br />
Neuromorphic engineering is an interdisciplinary subject that takes inspiration from [[biology]], [[physics]], [[mathematics]], [[computer science]], and [[electronic engineering]]<ref name=":2" /> to design artificial neural systems, such as [[Machine vision|vision systems]], head-eye systems, auditory processors, and autonomous robots, whose physical architecture and design principles are based on those of biological nervous systems.<ref name=":9">{{Cite journal | doi = 10.1155/2012/705483| title = Qualitative Functional Decomposition Analysis of Evolved Neuromorphic Flight Controllers| journal = Applied Computational Intelligence and Soft Computing| volume = 2012| pages = 1–21| year = 2012| last1 = Boddhu | first1 = S. K. | last2 = Gallagher | first2 = J. C. | doi-access = free}}</ref> It was developed by [[Carver Mead]]<ref name=":10">{{cite web|title=carver mead website|last1=Mead|first1=Carver|url=http://www.carvermead.caltech.edu/index.html|website=carvermead}}</ref> in the late 1980s.<br />
<br />
<br />
神经形态工程是以'''<font color="#ff8000">生物学Biology</font>'''、'''<font color="#ff8000">物理学Physics</font>'''、'''<font color="#ff8000">数学Mathematics</font>'''、'''<font color="#ff8000">计算机科学Computer science</font>'''和'''<font color="#ff8000">电子工程Electronic engineering<ref name=":2" /></font>'''等学科为基础，设计人工神经系统（如'''<font color="#ff8000">视觉系统Vision systems</font>'''、头眼系统、听觉处理器以及物理结构和设计原则都受启发于生物神经系统的自主机器人）的一门交叉学科。<ref name=":9" />20世纪80年代后期，卡弗·米德极大地推动了神经形态工程领域的发展。<ref name=":10" /><br />
<br />
==Neurological inspiration==<br />
==神经生物学启发==<br />
Neuromorphic engineering is set apart by the inspiration it takes from what we know about the structure and operations of the [[brain]]. Neuromorphic engineering translates what we know about the brain's function into computer systems. Work has mostly focused on replicating the analog nature of [[biological computation]] and the role of [[neuron]]s in [[cognition]].<br />
<br />
<br />
神经形态工程形成的灵感来源于现已知的'''<font color="#ff8000">大脑Brain</font>'''结构及运作机制知识，它将我们对大脑功能的了解用于研究和优化计算机系统。该领域工作主要集中于对'''<font color="#ff8000">生物计算Biological computation</font>'''模拟特性和'''<font color="#ff8000">神经元Neuron</font>'''在'''<font color="#ff8000">认知Cognition</font>'''中所发挥作用的复刻。<br />
<br />
The biological processes of neurons and their [[synapse]]s are dauntingly complex, and thus very difficult to artificially simulate. A key feature of biological brains is that all of the processing in neurons use analog chemical signals. This makes it hard to replicate brains in computers because the current generation of computers is completely digital. However, the characteristics of these parts can be abstracted into mathematical functions that closely capture the essence of the neuron's operations.<br />
<br />
神经元及其'''<font color="#ff8000">突触Synapse</font>'''的生理学过程极其复杂，难以进行人工模拟。神经元在所有处理过程中都使用化学模拟信号，这是大脑的一个关键生理学特征。这个特征大大增加了在计算机中复制大脑的难度，因为目前的计算机是完全数字化的。然而，部分特征可以抽象为数学函数，这些函数能够紧密捕捉神经元运行的本质。<br />
<br />
The goal of neuromorphic computing is not to perfectly mimic the brain and all of its functions, but instead to extract what is known of its structure and operations to be used in a practical computing system. No neuromorphic system will claim nor attempt to reproduce every element of neurons and synapses, but all adhere to the idea that computation is highly [[distributed processing|distributed]] throughout a series of small computing elements analogous to a neuron. While this sentiment is standard, researchers chase this goal with different methods.<ref name=":11">{{Cite journal | doi = 10.1088/1741-2560/13/5/051001| title = Large-scale neuromorphic computing systems| journal = Journal of Neural Engineering| volume = 13| pages = 1–15| year = 2016| last1 = Furber | first1 = Steve| issue = 5| pmid = 27529195| bibcode = 2016JNEng..13e1001F| doi-access = free}}</ref><br />
<br />
神经形态计算的目标不是完美地模拟大脑及其所有功能，而是利用已知的大脑结构和运转机制来研发或优化实际的计算系统。任何神经形态学系统都不会声称或试图复制神经元和突触中的每一个元素，但所有人都一致认可将计算高度'''<font color="#ff8000">分散Distribute</font>'''于一系列类似于神经元的小型计算元素的理念。研究人员用不同的方法来追求这一普遍目标。<ref name=":11" /><br />
<br />
<br />
==Examples==<br />
==代表性成果==<br />
As early as 2006, researchers at [[Georgia Tech]] published a field programmable neural array.<ref name=":12">{{Cite book|title = A field programmable neural array|last1 = Farquhar|first1 = Ethan|date = May 2006|journal = IEEE International Symposium on Circuits and Systems|pages = 4114–4117|last2 = Hasler|first2 = Paul.|doi = 10.1109/ISCAS.2006.1693534|isbn = 978-0-7803-9389-9|s2cid = 206966013}}</ref> This chip was the first in a line of increasingly complex arrays of floating gate transistors that allowed programmability of charge on the gates of [[MOSFET]]s to model the channel-ion characteristics of neurons in the brain and was one of the first cases of a silicon programmable array of neurons.<br />
<br />
早在2006年，佐治亚理工学院的研究人员就研发出了一种现场可编程神经阵列。<ref name=":12" />在此之后，一系列越来越复杂的浮栅晶体管阵列被成功研发出来，这些晶体管阵列可以通过在'''<font color="#ff8000">金属-氧化物半导体效应晶体管MOSFET</font>'''的栅极上编程来模拟大脑中神经元的离子通道特性，这也是以硅为主要材料的可编程神经元阵列的最早成功案例之一。<br />
<br />
In November 2011, a group of [[MIT]] researchers created a computer chip that mimics the analog, ion-based communication in a synapse between two neurons using 400 transistors and standard [[CMOS]] manufacturing techniques.<ref name=":13">{{cite web|title=MIT creates "brain chip"|url=http://www.extremetech.com/extreme/105067-mit-creates-brain-chip|access-date=4 December 2012}}</ref><ref name="Neuromorphic silicon paper">{{cite journal|title=Neuromorphic silicon neurons and large-scale neural networks: challenges and opportunities|doi=10.3389/fnins.2011.00108|pmid=21991244|pmc=3181466|volume=5|pages=108|journal=Frontiers in Neuroscience|year=2011|last1=Poon|first1=Chi-Sang|last2=Zhou|first2=Kuan|doi-access=free}}</ref><br />
<br />
2011年11月，麻省理工学院的一组研究人员研发出一种计算机芯片，该芯片上使用标准的'''<font color="#ff8000">互补金属氧化物半导体CMOS</font>'''制造技术集成了400个晶体管，能够模拟神经元间突触中基于离子的通讯。<ref name=":13" /><ref name="Neuromorphic silicon paper" /><br />
<br />
In June 2012, [[spintronic]] researchers at [[Purdue University]] presented a paper on the design of a neuromorphic chip using [[Spin valve|lateral spin valve]]s and [[memristor]]s. They argue that the architecture works similarly to neurons and can therefore be used to test methods of reproducing the brain's processing. In addition, these chips are significantly more energy-efficient than conventional ones.<ref name="Spin Devices Prop">{{Cite arXiv|title=Proposal For Neuromorphic Hardware Using Spin Devices|eprint=1206.3227|last1=Sharad|first1=Mrigank|last2=Augustine|first2=Charles|last3=Panagopoulos|first3=Georgios|last4=Roy|first4=Kaushik|class=cond-mat.dis-nn|year=2012}}</ref><br />
<br />
2012年6月，普渡大学的'''<font color="#ff8000">自旋电子学Spintronic</font>'''研究人员发表了一篇关于利用'''<font color="#ff8000">侧向自旋阀Lateral spin valves</font>'''和'''<font color="#ff8000">记忆电阻器Memristors</font>'''设计神经形态芯片的论文。他们认为，这种芯片结构的工作原理与神经元相似，因此可以用于大脑运行机制的复刻方法的测试。此外，这些芯片在能耗方面明显优于传统芯片。<ref name="Spin Devices Prop" /><br />
<br />
Research at [[HP Labs]] on Mott memristors has shown that while they can be non-[[Volatile memory|volatile]], the volatile behavior exhibited at temperatures significantly below the [[phase transition]] temperature can be exploited to fabricate a [[neuristor]],<ref name=":0" /> a biologically-inspired device that mimics behavior found in neurons.<ref name=":0">{{Cite journal | doi = 10.1038/nmat3510| pmid = 23241533| title = A scalable neuristor built with Mott memristors| journal = Nature Materials| volume = 12| issue = 2| pages = 114–7| year = 2012| last1 = Pickett | first1 = M. D. | last2 = Medeiros-Ribeiro | first2 = G. | last3 = Williams | first3 = R. S. | bibcode = 2013NatMa..12..114P| s2cid = 16271627| url = https://semanticscholar.org/paper/b6ba6f496ace2b947f111059663e76bb60e9efeb}}</ref> In September 2013, they presented models and simulations that show how the spiking behavior of these neuristors can be used to form the components required for a [[Turing machine]].<ref name=":14">{{cite journal|doi=10.1088/0957-4484/24/38/384002|title=Phase transitions enable computational universality in neuristor-based cellular automata|author1=Matthew D Pickett|author2=R Stanley Williams|name-list-style=amp|date=September 2013|publisher=IOP Publishing Ltd|journal=Nanotechnology|volume=24|issue=38|pmid=23999059|bibcode=2013Nanot..24L4002P|at=384002}}</ref><br />
<br />
'''<font color="#ff8000">惠普实验室HP labs</font>'''在莫特记忆电阻器上的研究表明，尽管它们可以是非'''<font color="#ff8000">易失性Volatile</font>'''的，但是在'''<font color="#ff8000">相变Phase transition</font>'''温度以下时表现出的易失性行为可以被用来制造'''<font color="#ff8000">类神经元电阻器Neuristor</font>'''(一种生物学启发的模仿神经元行为的硬件)<ref name=":0" />。2013年9月，他们通过模型和仿真展示了这些类神经元电阻器的脉冲行为如何产生'''<font color="#ff8000">图灵机Turing machine</font>'''的所需元素。<ref name=":14" /><br />
<br />
[[Neurogrid]], built by ''Brains in Silicon'' at [[Stanford University]],<ref name=":15">{{cite journal|last1=Boahen|first1=Kwabena|title=Neurogrid: A Mixed-Analog-Digital Multichip System for Large-Scale Neural Simulations|journal=Proceedings of the IEEE|date=24 April 2014|volume=102|issue=5|pages=699–716|doi=10.1109/JPROC.2014.2313565|s2cid=17176371}}</ref> is an example of hardware designed using neuromorphic engineering principles. The circuit board is composed of 16 custom-designed chips, referred to as NeuroCores. Each NeuroCore's analog circuitry is designed to emulate neural elements for 65536 neurons, maximizing energy efficiency. The emulated neurons are connected using digital circuitry designed to maximize spiking throughput.<ref name=":16">{{cite journal|doi=10.1038/503022a|pmid = 24201264|title = Neuroelectronics: Smart connections|journal = Nature|volume = 503|issue = 7474|pages = 22–4|year = 2013|last1 = Waldrop|first1 = M. Mitchell|bibcode = 2013Natur.503...22W|doi-access = free}}</ref><ref name=":17">{{cite journal|doi=10.1109/JPROC.2014.2313565|title = Neurogrid: A Mixed-Analog-Digital Multichip System for Large-Scale Neural Simulations|journal = Proceedings of the IEEE|volume = 102|issue = 5|pages = 699–716|year = 2014|last1 = Benjamin|first1 = Ben Varkey|last2 = Peiran Gao|last3 = McQuinn|first3 = Emmett|last4 = Choudhary|first4 = Swadesh|last5 = Chandrasekaran|first5 = Anand R.|last6 = Bussat|first6 = Jean-Marie|last7 = Alvarez-Icaza|first7 = Rodrigo|last8 = Arthur|first8 = John V.|last9 = Merolla|first9 = Paul A.|last10 = Boahen|first10 = Kwabena|s2cid = 17176371}}</ref><br />
<br />
'''<font color="#ff8000">神经栅格Neurogrid<ref name=":15" /></font>'''是由斯坦福大学Brains in Silicon公司研发的、使用神经形态工程原理设计的硬件。该电路板由16个定制设计的芯片组成(NeuroCores)。在设计中，每个NeuroCore芯片的模拟电路对65536个神经元的神经元素进行模拟，以最大限度地提高能量效率。模拟出的神经元通过设计的数字电路连接，以最大化脉冲吞吐量。<ref name=":16" /><ref name=":17" /><br />
<br />
A research project with implications for neuromorphic engineering is the [[Human Brain Project]] that is attempting to simulate a complete human brain in a supercomputer using biological data. It is made up of a group of researchers in neuroscience, medicine, and computing.<ref name=":18">{{cite web|title=Involved Organizations|url=http://www.humanbrainproject.eu/partners.html|access-date=22 February 2013|url-status=dead|archive-url=https://web.archive.org/web/20130302142627/http://www.humanbrainproject.eu/partners.html|archive-date=2 March 2013}}</ref> [[Henry Markram]], the project's co-director, has stated that the project proposes to establish a foundation to explore and understand the brain and its diseases, and to use that knowledge to build new computing technologies. The three primary goals of the project are to better understand how the pieces of the brain fit and work together, to understand how to objectively diagnose and treat brain diseases, and to use the understanding of the human brain to develop neuromorphic computers. That the simulation of a complete human brain will require a supercomputer a thousand times more powerful than today's encourages the current focus on neuromorphic computers.<ref name=":19">{{cite web|title=Human Brain Project|url=http://www.humanbrainproject.eu|access-date=22 February 2013}}</ref> $1.3 billion has been allocated to the project by The [[European Commission]].<ref name=":20">{{cite web|title=The Human Brain Project and Recruiting More Cyberwarriors|url=http://www.marketplace.org/topics/tech/human-brain-project-and-recruiting-more-cyberwarriors|access-date=22 February 2013|date=January 29, 2013}}</ref><br />
<br />
'''<font color="#ff8000">人类大脑计划Human Brain Project</font>'''对神经形态工程具有较大影响，其主要任务是尝试用生物数据在超级计算机中模拟完整的人脑。人类大脑计划由神经科学、医学和计算机科学背景的研究人员组成。<ref name=":18" />该项目的联合主管亨利•马克拉姆(Henry Markram)表示，人类大脑计划的目的是建立一个探索和了解脑科学和脑疾病知识的基础，并利用这些知识来构建更先进的计算机技术。这个项目的三个主要目标分别是: 更好地理解大脑的各个部分是如何相互配合协同工作的; 理解如何客观地诊断和治疗脑部疾病; 以及利用对人类大脑的理解来开发神经形态计算机。模拟一个完整的人类大脑需要一台比现在强大一千倍的超级计算机，这不断激发着对神经形态计算机领域的研究兴趣。<ref name=":19" />欧盟委员会已经向人类大脑计划拨款13亿美元。<ref name=":20" /><br />
<br />
Other research with implications for neuromorphic engineering involves the [[BRAIN Initiative]]<ref name="economist">[https://www.economist.com/news/science-and-technology/21582495-computers-will-help-people-understand-brains-better-and-understanding-brains Neuromorphic computing: The machine of a new soul], The Economist, 2013-08-03</ref> and the [[TrueNorth]] chip from [[IBM]].<ref name=":21">{{cite journal|last1=Modha|first1=Dharmendra|title=A million spiking-neuron integrated circuit with a scalable communication network and interface|journal=Science|date=Aug 2014|volume=345|issue=6197|pages=668–673|doi=10.1126/science.1254642|pmid=25104385|bibcode=2014Sci...345..668M|s2cid=12706847}}</ref> Neuromorphic devices have also been demonstrated using nanocrystals, nanowires, and conducting polymers.<ref name=":22">{{Cite web|url=http://jessamynfairfield.com/wp-content/uploads/2017/03/PWMar17Fairfield.pdf|title=Smarter Machines|last=Fairfield|first=Jessamyn|date=March 1, 2017}}</ref><br />
<br />
其他与神经形态工程有关的研究还包括'''<font color="#ff8000">脑计划BRAIN initiative</font>'''，<ref name="economist" />和IBM研发的'''<font color="#ff8000">TrueNorth</font>'''芯片。<ref name=":21" />使用纳米晶体、纳米线和导电聚合物也能够用于制造神经形态学硬件。<ref name=":22" /><br />
<br />
[[Intel]] unveiled its neuromorphic research chip, called “[[Intel Loihi|Loihi]]”, in October 2017. The chip uses an asynchronous [[spiking neural network]] (SNN) to implement adaptive self-modifying event-driven fine-grained parallel computations used to implement learning and inference with high efficiency.<ref name=":23">{{cite journal |last1=Davies |first1=Mike |title=Loihi: A Neuromorphic Manycore Processor with On-Chip Learning |journal=IEEE Micro |date=January 16, 2018 |volume=38 |issue=1 |pages=82–99 |display-authors=etal|doi=10.1109/MM.2018.112130359 |s2cid=3608458 }}</ref><ref name="Morris2017">{{cite web |last1=Morris |first1=John |title=Why Intel built a neuromorphic chip |url=https://www.zdnet.com/article/why-intel-built-a-neuromorphic-chip/ |website=ZDNet |access-date=17 August 2018 |language=en}}</ref><br />
<br />
2017年10月，英特尔发布了神经形态芯片'''<font color="#ff8000">Loihi</font>'''。该芯片采用异步'''<font color="#ff8000">脉冲神经网络Spiking neural network</font>'''实现了自适应、自修改、事件驱动的细粒度并行计算，实现了高效的学习和推理。<ref name=":23" /><ref name="Morris2017" /><br />
<br />
[[IMEC]], a Belgium-based nanoelectronics research center, demonstrated the world's first self-learning neuromorphic chip. The brain-inspired chip, based on OxRAM technology, has the capability of self-learning and has been demonstrated to have the ability to compose music.<ref name=":24">{{cite web |title=Imec demonstrates self-learning neuromorphic chip that composes music |url=https://www.imec-int.com/en/articles/imec-demonstrates-self-learning-neuromorphic-chip-that-composes-music |website=IMEC International |access-date=1 October 2019}}</ref> IMEC released the 30-second tune composed by the prototype. The chip was sequentially loaded with songs in the same time signature and style. The songs were old Belgian and French flute minuets, from which the chip learned the rules at play and then applied them.<ref name=":25">{{cite web|last1=Bourzac|first1=Katherine|title=A Neuromorphic Chip That Makes Music|url=https://spectrum.ieee.org/a-neuromorphic-chip-that-makes-music|url-status=live|access-date=1 October 2019|website=IEEE Spectrum|date=May 23, 2017}}</ref><br />
<br />
比利时的'''<font color="#ff8000">微电子研究中心IMEC</font>'''研发了世界上首个自学习神经形态芯片。这种基于'''<font color="#32CD32"> OxRAM(filamentary-oxide-based resistive memory technology)</font>''' 技术的大脑启发芯片具有自学习能力，并且已被证明具有创作音乐的能力。<ref name=":24" />IMEC发布了由芯片原型机谱写的30秒旋律。向芯片加载一系列特征、风格统一的歌曲（古代比利时和法国长笛小步舞曲），芯片从中学习相关规则并将其应用于创作。<ref name=":25" /><br />
<br />
[[Blue Brain Project|The Blue Brain Project]], led by Henry Markram, aims to build biologically detailed digital reconstructions and simulations of the mouse brain. The Blue Brain Project has created in silico models of rodent brains, while attempting to replicate as many details about its biology as possible. The supercomputer-based simulations offer new perspectives on understanding the structure and functions of the brain.<br />
<br />
由亨利·马克拉姆领导的'''<font color="#ff8000">蓝脑计划The Blue Brain Project</font>'''旨在建立小鼠大脑生理学细节的数字重建和模拟。蓝脑计划已经建立了啮齿动物大脑的电子模型，同时进行着尽可能多地复制其生理学细节的尝试。基于超级计算机的模拟为理解大脑的结构和功能提供了新的视角。<br />
<br />
The European Union funded a series of projects at the University of Heidelberg, which led to the development of <br />
[[BrainScaleS]] (brain-inspired multiscale computation in neuromorphic hybrid systems), a hybrid analog [[neuromorphic]] supercomputer located at Heidelberg University, Germany. It was developed as part of the [[Human Brain Project]] neuromorphic computing platform and is the complement to the [[SpiNNaker]] supercomputer (which is based on digital technology). The architecture used in BrainScaleS mimics biological neurons and their connections on a physical level; additionally, since the components are made of silicon, these model neurons operate on average 864 times (24 hours of real time is 100 seconds in the machine simulation) that of their biological counterparts.<ref name=":26">{{Cite web|date=2016-03-21|title=Beyond von Neumann, Neuromorphic Computing Steadily Advances|url=https://www.hpcwire.com/2016/03/21/lacking-breakthrough-neuromorphic-computing-steadily-advance/|access-date=2021-10-08|website=HPCwire|language=en-US}}</ref><br />
<br />
欧盟资助了海德堡大学一系列促进BrainScaleS(神经形态混合系统中受大脑启发的多尺度计算)发展的项目，这是一台位于德国海德堡大学的混合模拟'''<font color="#ff8000">神经形态Neuromorphic</font>'''超级计算机。它是作为人类大脑计划中神经形态计算平台的一部分而开发的，是'''<font color="#ff8000">SpiNNaker</font>'''超级计算机(基于数字技术)的补充。BrainScaleS中使用的体系架构模拟了生物神经元及其在物理层面上的连接；此外，由于这些组件是由硅制成的，这些模型神经元平均运行速度是生物神经元的864倍，这意味着在机器模拟中，24小时的实时时间仅为100秒。<ref name=":26" /><br />
<br />
===Neuromorphic sensors===<br />
===神经形态传感器===<br />
The concept of neuromorphic systems can be extended to sensors (not just to computation). An example of this applied to detecting [[light]] is the [[retinomorphic sensor]] or, when employed in an array, the [[event camera]].<br />
<br />
神经形态系统的概念可以扩展到传感器(而不仅仅是计算单元)。一个用于检测光线的例子是'''<font color="#ff8000">类视网膜传感器Retinomorphic sensor</font>'''，或者'''<font color="#ff8000">事件摄像机Event camera</font>阵列。'''<br />
<br />
<br />
==Ethical considerations ==<br />
==伦理问题==<br />
While the interdisciplinary concept of neuromorphic engineering is relatively new, many of the same ethical considerations apply to neuromorphic systems as apply to [[human-like machines]] and [[artificial intelligence]] in general. However, the fact that neuromorphic systems are designed to mimic a [[human brain]] gives rise to unique ethical questions surrounding their usage.<br />
<br />
虽然神经形态工程这个跨学科概念相对较新，但许多适用于'''<font color="#ff8000">类人机器Human-like machines</font>'''和人工智能的伦理考虑大体上也适用于神经形态系统。另一方面，神经形态系统是为了模仿人类大脑而设计，这一事实引起了一些针对性的独特伦理问题。<br />
<br />
However, the practical debate is that neuromorphic hardware as well as artificial "neural networks" are immensely simplified models of how the brain operates or processes information at a much lower [[complex system|complexity in terms of size and functional technology]] and a much more regular structure in terms of [[brain connectivity|connectivity]]. Comparing [[neuromorphic chip]]s to the brain is a very crude comparison similar to comparing a plane to a bird just because they both have wings and a tail. The fact is that neural cognitive systems are many orders of magnitude more [[Energy efficiency (physics)|energy-]] and compute-efficient than current state-of-the-art AI and neuromorphic engineering is an attempt to narrow this gap by inspiring from the brain's mechanism just like many engineering designs have [[bioengineering|bio-inspired features]].<br />
<br />
然而，这种争论的实际情况是，神经形态硬件和人工“神经网络”是对大脑运作或处理信息过程的极其简化的模型，在大小和功能技术方面的'''<font color="#ff8000">复杂性Complexity</font>'''相比而言比较低，在'''<font color="#ff8000">连接性Connectivity</font>'''方面也具有更加规则的结构。将'''<font color="#ff8000">神经形态芯片Neuromorphic chips</font>'''与大脑进行比较是一种非常粗糙的比较，类似于仅仅因为飞机有翅膀和尾巴就将它与鸟进行比较。事实上，当前最先进的人工智能在'''<font color="#ff8000">能耗效率Energy-efficiency</font>'''和计算效率方面距离人脑神经认知系统仍有较大差距，而神经形态工程只是一种通过从大脑机制中获得灵感来缩小这种差距的尝试，就像许多工程设计中都具有'''<font color="#ff8000">生物启发的特征Bio-inspired features</font>'''一样。<br />
<br />
===Democratic concerns===<br />
===公众担忧===<br />
Significant ethical limitations may be placed on neuromorphic engineering due to public perception.<ref name=":27">{{Cite report|url=https://ai100.stanford.edu/sites/g/files/sbiybj9861/f/ai_100_report_0831fnl.pdf|title=Artificial Intelligence and Life in 2030|author=2015 Study Panel|date=September 2016|work=One Hundred Year Study on Artificial Intelligence (AI100)|publisher=Stanford University}}</ref> Special [[Eurobarometer]] 382: Public Attitudes Towards Robots, a survey conducted by the European Commission, found that 60% of [[European Union]] citizens wanted a ban of robots in the care of children, the elderly, or the disabled. Furthermore, 34% were in favor of a ban on robots in education, 27% in healthcare, and 20% in leisure. The European Commission classifies these areas as notably “human.” The report cites increased public concern with robots that are able to mimic or replicate human functions. Neuromorphic engineering, by definition, is designed to replicate the function of the human brain.<ref name=":1">{{Cite web|url=http://ec.europa.eu/commfrontoffice/publicopinion/archives/ebs/ebs_382_en.pdf|title=Special Eurobarometer 382: Public Attitudes Towards Robots|last=European Commission|date=September 2012|website=European Commission}}</ref><br />
<br />
由于公众认知的相关忧虑，神经形态工程学可能会受到严重的伦理限制。<ref name=":27" />欧盟委员会进行的一项调查发现，60% 的欧盟公民希望禁止机器人参与照顾儿童、老人或残疾人的工作。此外，34% 的人支持禁止机器人用于教育，27% 的人支持禁止机器人用于医疗保健，20% 的人支持禁止机器人用于休闲。欧盟委员会将以上领域划入“人类”范畴。报告指出，公众越来越关注能够模仿或复制人类功能的机器人。神经形态工程，顾名思义，是为了复制人脑的功能而设计的。<ref name=":1" /><br />
<br />
The democratic concerns surrounding neuromorphic engineering are likely to become even more profound in the future. The European Commission found that EU citizens between the ages of 15 and 24 are more likely to think of robots as human-like (as opposed to instrument-like) than EU citizens over the age of 55. When presented an image of a robot that had been defined as human-like, 75% of EU citizens aged 15–24 said it corresponded with the idea they had of robots while only 57% of EU citizens over the age of 55 responded the same way. The human-like nature of neuromorphic systems, therefore, could place them in the categories of robots many EU citizens would like to see banned in the future.<ref name=":1" /><br />
<br />
围绕神经形态工程的公众担忧可能在未来变得更加深刻。欧盟委员会发现，相比于55岁以上的欧盟公民，15至24岁的欧盟公民更有可能认为机器人像人(而不是像仪器)。当看到一张“类人”机器人的图片时，年龄在15岁至24岁之间的欧盟公民中有75% 的人表示这符合他们对机器人的想法，而55岁以上的欧盟公民中只有57% 的人有同样的反应。因此，神经形态系统可能因为其类似人类的特性而被归入许多欧盟公民希望在未来禁止使用的机器人类别。<ref name=":1" /><br />
<br />
=== Personhood===<br />
=== 人格问题===<br />
As neuromorphic systems have become increasingly advanced, some scholars{{who|date=August 2021}} have advocated for granting [[personhood]] rights to these systems. If the brain is what grants humans their personhood, to what extent does a neuromorphic system have to mimic the human brain to be granted personhood rights? Critics of technology development in the [[Human Brain Project]], which aims to advance brain-inspired computing, have argued that advancement in neuromorphic computing could lead to machine consciousness or personhood.<ref name=":28">{{Cite journal|last=Aicardi|first=Christine|date=September 2018|title=Accompanying technology development in the Human Brain Project: From foresight to ethics management|journal=Futures|volume=102|pages=114–124|doi=10.1016/j.futures.2018.01.005|doi-access=free}}</ref> If these systems are to be treated as people, critics argue, then many tasks humans perform using neuromorphic systems, including the act of termination of neuromorphic systems, may be morally impermissible as these acts would violate the autonomy of the neuromorphic systems.<ref name=":29">{{Cite journal|last=Lim|first=Daniel|date=2014-06-01|title=Brain simulation and personhood: a concern with the Human Brain Project|journal=Ethics and Information Technology|language=en|volume=16|issue=2|pages=77–89|doi=10.1007/s10676-013-9330-5|s2cid=17415814|issn=1572-8439}}</ref><br />
<br />
随着神经形态系统的日益发展，一些学者主张赋予这些系统'''<font color="ff8000">人格Personhood</font>'''权。如果是大脑赋予了人类人格，那么在多大程度上模仿人类大脑的神经形态系统才能被赋予人格权利？“人类大脑计划”旨在推进以大脑为灵感的计算机技术发展，该计划的批评者认为，神经形态计算机技术的进步可能导致机器意识或人格的形成。<ref name=":28" />这些批评者认为，如果这些系统被当作人来对待，那么人类使用神经形态系统执行任务（包括终止神经形态系统）的行为，在道德上就可能是不被允许的，因为这些行为将违反神经形态系统的自主性。<ref name=":29" /><br />
<br />
<br />
==Dual use (military applications)==<br />
==军民两用技术（军事应用）==<br />
The [[Joint Artificial Intelligence Center]], a branch of the U.S. military, is a center dedicated to the procurement and implementation of AI software and neuromorphic hardware for combat use. Specific applications include smart headsets/goggles and robots. JAIC intends to rely heavily on neuromorphic technology to connect "every fighter every shooter" within a network of neuromorphic-enabled units.<br />
<br />
'''<font color="#ff8000">联合人工智能中心The Joint Artificial Intelligence Center</font>'''(JAIC),是美国军队的一个分支，专门从事采购和实施用于战斗的人工智能软件和神经形态硬件。具体应用包括智能耳机/护目镜和机器人。JAIC打算高度依赖神经形态技术，'''<font color="#32CD32">使用神经形态技术来连接神经形态单位网络中的“每个战士每个射手”</font>'''。<br />
<br />
<br />
==Legal considerations==<br />
==法律问题==<br />
Skeptics have argued that there is no way to apply the electronic personhood, the concept of personhood that would apply to neuromorphic technology, legally. In a letter signed by 285 experts in law, robotics, medicine, and ethics opposing a European Commission proposal to recognize “smart robots” as legal persons, the authors write, “A legal status for a robot can’t derive from the [[Natural person|Natural Person]] model, since the robot would then hold [[human rights]], such as the right to dignity, the right to its integrity, the right to remuneration or the right to citizenship, thus directly confronting the Human rights. This would be in contradiction with the [[Charter of Fundamental Rights of the European Union]] and the [[Convention for the Protection of Human Rights and Fundamental Freedoms]].”<ref name=":30">{{Cite web|url=http://www.robotics-openletter.eu/|title=Robotics Openletter {{!}} Open letter to the European Commission|language=fr-FR|access-date=2019-05-10}}</ref><br />
<br />
质疑者者认为，在法律上没有办法应用能够适用于神经形态技术的电子人格。在一封由285名法律、机器人技术、医学和伦理学专家的联名信中，作者们反对欧盟委员会提出的承认“智能机器人”为法人的提议。他们写道，“机器人的法律地位无法从'''<font color="#ff8000">自然人Natural Person</font>'''模型中推导出来，因为机器人将被赋予'''<font color="#ff8000">人权Human rights</font>'''，如尊严权、完整权、报酬权或公民权，从而直接面临人权问题。这将有悖于《欧联基本权利宪章》和《欧洲保障人权和根本自由公约》”。<ref name=":30" /><br />
<br />
===Ownership and property rights ===<br />
===所有权及财产权问题===<br />
There is significant legal debate around property rights and artificial intelligence. In ''Acohs Pty Ltd v. Ucorp Pty Ltd'', Justice Christopher Jessup of the [[Federal Court of Australia]] found that the [[source code]] for [[Material safety data sheets|Material Safety Data Sheets]] could not be [[Copyright law of Australia|copyrighted]] as it was generated by a [[software interface]] rather than a human author.<ref name=":31">{{Cite web|url=http://www.lavan.com.au/advice/intellectual_property_technology/copyright_in_source_code_and_digital_products|title=Copyright in source code and digital products|last=Lavan|website=Lavan|language=en|access-date=2019-05-10}}</ref> The same question may apply to neuromorphic systems: if a neuromorphic system successfully mimics a human brain and produces a piece of original work, who, if anyone, should be able to claim ownership of the work?<ref name=":32">{{cite journal |last1=Eshraghian|first1=Jason K. |title=Human Ownership of Artificial Creativity |journal=Nature Machine Intelligence |date=9 March 2020 |volume=2 |pages=157–160 |doi=10.1038/s42256-020-0161-x}}</ref><br />
<br />
法律界围绕财产权和人工智能有着重大争论。在Acohs Pty有限公司诉Ucorp Pty有限公司一案中，澳大利亚联邦法院的克里斯托弗·杰瑟普法官发现，版权保护不适用于材料安全数据表的'''<font color="#ff8000">源代码Source code</font>'''，因为它是由'''<font color="#ff8000">软件界面Software interface</font>'''生成而非人类工作者生成的。<ref name=":31" />同样的问题可能也适用于神经形态系统: 如果一个神经形态系统成功地模仿了人类的大脑，并产生了一部原创作品，那么该如何确认这部作品的所有权归属？<ref name=":32" /><br />
<br />
<br />
==Neuromemristive systems==<br />
==神经记忆电阻系统==<br />
Neuromemristive systems are a subclass of neuromorphic computing systems that focus on the use of [[memristors]] to implement [[neuroplasticity]]. While neuromorphic engineering focuses on mimicking biological behavior, neuromemristive systems focus on abstraction.<ref name=":33">{{Cite web|url=https://digitalops.sandia.gov/Mediasite/Play/a10cf6ceb55d47608bb8326dd00e46611d|title=002.08 N.I.C.E. Workshop 2014: Towards Intelligent Computing with Neuromemristive Circuits and Systems - Feb. 2014|website=digitalops.sandia.gov|access-date=2019-08-26}}</ref> For example, a neuromemristive system may replace the details of a [[Cerebral cortex|cortical]] microcircuit's behavior with an abstract neural network model.<ref name=":34">C. Merkel and D. Kudithipudi, "Neuromemristive extreme learning machines for pattern classification," ISVLSI, 2014.</ref><br />
<br />
神经记忆电阻系统是神经形态计算系统的一个亚类，主要研究利用'''<font color="ff8000">记忆电阻器Memristors</font>'''实现'''<font color="ff8000">神经可塑性Neuroplasticity</font>'''。神经形态工程的重点是模拟生物行为，而神经记忆电阻系统的重点是提取。<ref name=":33" />举个例子，一个神经记忆系统可能用抽象的神经网络模型替代'''<font color="ff8000">皮层Cortical</font>'''微电路的行为细节。<ref name=":34" /><br />
<br />
There exist several neuron inspired threshold logic functions<ref name="Maan 1–13" /> implemented with memristors that have applications in high level [[pattern recognition]] applications. Some of the applications reported recently include [[speech recognition]],<ref name=":35">{{Cite journal|title = Memristor pattern recogniser: isolated speech word recognition|journal = Electronics Letters|pages = 1370–1372|volume = 51|issue = 17|doi = 10.1049/el.2015.1428|first1 = A.K.|last1 = Maan|first2 = A.P.|last2 = James|first3 = S.|last3 = Dimitrijev|year = 2015|bibcode = 2015ElL....51.1370M|hdl = 10072/140989|s2cid = 61454815|url = https://semanticscholar.org/paper/48d3ab11ec6e213b62f11eedcfb7b7febb058674|hdl-access = free}}</ref> [[face recognition]]<ref name=":36">{{Cite journal|title = Memristive Threshold Logic Face Recognition|journal = Procedia Computer Science|date = 2014-01-01|pages = 98–103|volume = 41|series = 5th Annual International Conference on Biologically Inspired Cognitive Architectures, 2014 BICA|doi = 10.1016/j.procs.2014.11.090|first1 = Akshay Kumar|last1 = Maan|first2 = Dinesh S.|last2 = Kumar|first3 = Alex Pappachen|last3 = James|doi-access = free}}</ref> and [[object recognition]].<ref name=":37">{{Cite journal|title = Memristive Threshold Logic Circuit Design of Fast Moving Object Detection|journal = IEEE Transactions on Very Large Scale Integration (VLSI) Systems|date = 2015-10-01|issn = 1063-8210|pages = 2337–2341|volume = 23|issue = 10|doi = 10.1109/TVLSI.2014.2359801|first1 = A.K.|last1 = Maan|first2 = D.S.|last2 = Kumar|first3 = S.|last3 = Sugathan|first4 = A.P.|last4 = James|arxiv = 1410.1267|s2cid = 9647290}}</ref> They also find applications in replacing conventional digital logic gates.<ref name=":38">{{Cite journal|title = Resistive Threshold Logic|journal = IEEE Transactions on Very Large Scale Integration (VLSI) Systems|date = 2014-01-01|issn = 1063-8210|pages = 190–195|volume = 22|issue = 1|doi = 10.1109/TVLSI.2012.2232946|first1 = A.P.|last1 = James|first2 = L.R.V.J.|last2 = Francis|first3 = D.S.|last3 = Kumar|arxiv = 1308.0090|s2cid = 7357110}}</ref><ref name=":39">{{Cite journal|title = Threshold Logic Computing: Memristive-CMOS Circuits for Fast Fourier Transform and Vedic Multiplication|journal = IEEE Transactions on Very Large Scale Integration (VLSI) Systems|date = 2015-11-01|issn = 1063-8210|pages = 2690–2694|volume = 23|issue = 11|doi = 10.1109/TVLSI.2014.2371857|first1 = A.P.|last1 = James|first2 = D.S.|last2 = Kumar|first3 = A.|last3 = Ajayan|arxiv = 1411.5255|s2cid = 6076956}}</ref><br />
<br />
受神经元启发、使用记忆电阻器实现的阈值逻辑函数<ref name="Maan 1–13" />在高级'''<font color="#ff8000">模式识别Pattern recognition</font>'''中有着广泛的应用，最近报道中其应用包括'''<font color="#ff8000">语音识别Speech recognition<ref name=":35" /></font>'''、'''<font color="#ff8000">人脸识别Face recognition<ref name=":36" /></font>'''和'''<font color="#ff8000">物体识别Object recognition<ref name=":37" /></font>'''。阈值逻辑函数还可以用来取代传统的数字逻辑门。<ref name=":38" /><ref name=":39" /><br />
<br />
For ideal passive memristive circuits there is an exact equation (Caravelli-Traversa-[[Di Ventra]] equation) for the internal memory of the circuit:<ref name=":40">{{cite journal |last=Caravelli |display-authors=etal|arxiv=1608.08651 |title=The complex dynamics of memristive circuits: analytical results and universal slow relaxation |year=2017 |doi=10.1103/PhysRevE.95.022140 |pmid= 28297937 |volume=95 |issue= 2 |pages= 022140 |journal=Physical Review E|bibcode=2017PhRvE..95b2140C |s2cid=6758362}}</ref><br />
<br />
对于理想的无源记忆电路，电路的内部记忆可以用精确的方程(Caravelli-Traversa-Di Ventra方程) 来描述:<ref name=":40" /><br />
<br />
:<math> \frac{d}{dt} \vec{W} = \alpha \vec{W}-\frac{1}{\beta} (I+\xi \Omega W)^{-1} \Omega \vec S </math><br />
<br />
as a function of the properties of the physical memristive network and the external sources. In the equation above, <math>\alpha</math> is the "forgetting" time scale constant, <math> \xi=r-1</math> and <math>r=\frac{R_\text{off}}{R_\text{on}}</math> is the ratio of ''off'' and ''on'' values of the limit resistances of the memristors, <math> \vec S </math> is the vector of the sources of the circuit and <math>\Omega</math> is a projector on the fundamental loops of the circuit. The constant <math>\beta</math> has the dimension of a voltage and is associated to the properties of the [[memristor]]; its physical origin is the charge mobility in the conductor. The diagonal matrix and vector <math>W=\operatorname{diag}(\vec W)</math> and <math>\vec W</math> respectively, are instead the internal value of the memristors, with values between 0 and 1. This equation thus requires adding extra constraints on the memory values in order to be reliable.<br />
<br />
Caravelli-Traversa-Di Ventra方程是描述物理记忆网络和外部源性质的函数。在上述方程中，<math>\alpha</math>是“遗忘”时间尺度常数，<math>\xi=r-1</math>，<math>r =\frac{R\text_{off}}{R_\text{on}}</math>是记忆电阻器off状态和on状态极限电阻值之比，<math>\vec S</math>是电路源的矢量，<math>\Omega</math>是电路基本环路的投影。常数<math>\beta</math>具有电压的量纲，与记忆电阻器的特性有关；其物理原型是导体中的电荷迁移率。对角矩阵和向量 <math>W=\operatorname{diag}(\vec W)</math>和<nowiki><math>\vec W</math></nowiki>'''<font color="#32CD32">是记忆电阻器的内阻</font>'''，值在0到1之间。因此，这个等式需要在'''<font color="32CD32">内存值</font>'''上添加额外约束以保证可靠性。<br />
<br />
<br />
==See also==<br />
==相关词条==<br />
{{Columns-list|colwidth=18em|<br />
* [[AI accelerator (computer hardware)|AI accelerator]]<br />
* [[Artificial brain]]<br />
* [[Biomorphic]]<br />
* [[Cognitive computer]]<br />
* [[Computation and Neural Systems]]<br />
* [[Differentiable programming]]<br />
* [[Event camera]]<br />
* [[Neurorobotics]]<br />
* [[Optical flow sensor]]<br />
* [[Physical neural network]]<br />
* [[SpiNNaker]]<br />
* [[SyNAPSE]]<br />
* [[Retinomorphic sensor]]<br />
* [[Vision chip]]<br />
* [[Vision processing unit]]<br />
* [[Zeroth (software)]]<br />
* [[Hardware for artificial intelligence]]<br />
}}<br />
<br />
{{Portal bar|Electronics}}<br />
<br />
== References==<br />
{{Reflist|40em}}<br />
<br />
==External links==<br />
<!--======================== {{No more links}} ============================<br />
| PLEASE BE CAUTIOUS IN ADDING MORE LINKS TO THIS ARTICLE. Wikipedia |<br />
| is not a collection of links nor should it be used for advertising. |<br />
| |<br />
| Excessive or inappropriate links WILL BE DELETED. |<br />
| See [[Wikipedia:External links]] & [[Wikipedia:Spam]] for details. |<br />
| |<br />
| If there are already plentiful links, please propose additions or |<br />
| replacements on this article's discussion page, or submit your link |<br />
| to the relevant category at the Open Directory Project (dmoz.org) |<br />
| and link back to that category using the {{dmoz}} template. |<br />
======================={{No more links}}=============================--><br />
{{Commons category|Neuromorphic Engineering}}<br />
*[https://web.archive.org/web/20150727034331/http://ine-web.org/workshops/workshops-overview Telluride Neuromorphic Engineering Workshop]<br />
*[https://archive.today/20130115190057/http://capocaccia.ethz.ch/ CapoCaccia Cognitive Neuromorphic Engineering Workshop]<br />
*[https://web.archive.org/web/20190716132350/http://www.ine-web.org/ Institute of Neuromorphic Engineering]<br />
*[http://www.ine-news.org/ INE news site].<br />
*[http://www.frontiersin.org/neuromorphic_engineering Frontiers in Neuromorphic Engineering Journal]<br />
*[http://www.cns.caltech.edu/ Computation and Neural Systems] department at the [[California Institute of Technology]].<br />
*[http://www.humanbrainproject.eu/ Human Brain Project official site]<br />
*[https://www.the-scientist.com/features/building-a-silicon-brain-65738 Building a Silicon Brain:] Computer chips based on biological neurons may help simulate larger and more-complex brain models. May 1, 2019. SANDEEP RAVINDRAN<br />
<br />
<br />
* Telluride Neuromorphic Engineering Workshop<br />
*CapoCaccia Cognitive Neuromorphic Engineering Workshop<br />
*Institute of Neuromorphic Engineering<br />
*INE news site.<br />
*Frontiers in Neuromorphic Engineering Journal<br />
*Computation and Neural Systems department at the California Institute of Technology.<br />
*Human Brain Project official site<br />
* Building a Silicon Brain: Computer chips based on biological neurons may help simulate larger and more-complex brain models. May 1, 2019. SANDEEP RAVINDRAN<br />
<br />
=<nowiki>外部链接</nowiki>= <br />
*碲化物神经形态工程工作室<br />
*CapoCaccia认知神经形态工程工作室<br />
*神经形态工程研究所<br />
*INE新闻站点。<br />
*《神经形态工程学前沿》<br />
*加州理工学院计算与神经系统系。<br />
*人脑项目官方网站<br />
*打造硅脑: 基于生物神经元的计算机芯片可能有助于模拟更大、更复杂的大脑模型。2019年5月1日。SANDEEP RAVINDRAN.<br />
<br />
{{Differentiable computing}}<br />
{{Authority control}}<br />
<br />
[[Category:Electrical engineering]]<br />
[[Category:Neuroscience]]<br />
[[分类:AI accelerators|*]]<br />
[[Category:Artificial intelligence]]<br />
[[Category:Robotics]]<br />
<br />
Category:Electrical engineering<br />
Category:Neuroscience<br />
*<br />
Category:Artificial intelligence<br />
Category:Robotics<br />
<br />
类别: 电气工程类别: 神经科学 * 类别: 人工智能类别: 机器人<br />
<br />
<noinclude><br />
<br />
<small>This page was moved from [[wikipedia:en:Neuromorphic engineering]]. Its edit history can be viewed at [[神经形态计算/edithistory]]</small></noinclude><br />
<br />
[[Category:待整理页面]]<br />
<br />
<references /></div>Shenky20https://wiki.swarma.org/index.php?title=%E7%A5%9E%E7%BB%8F%E5%BD%A2%E6%80%81%E8%AE%A1%E7%AE%97&diff=30750神经形态计算2022-04-28T11:49:27Z<p>Shenky20：/* 神经记忆电阻系统 */</p>
<hr />
<div><br />
此词条由神经动力学模型读书会词条梳理志愿者Shenky20翻译审校，未经专家审校，带来阅读不便，请见谅。<br />
<br />
{{Use American English|date = January 2019}}<br />
{{Use mdy dates|date = January 2019}}<br />
<br />
'''Neuromorphic engineering''', also known as '''neuromorphic computing''',<ref name=":3">{{Cite journal | doi = 10.1145/2601069| title = Neuromorphic computing gets ready for the (really) big time| journal = [[Communications of the ACM]]| volume = 57| issue = 6| year = 2014| last1 = Monroe | first1 = D. | pages = 13–15| s2cid = 20051102}}</ref><ref name=":4">{{Cite journal | doi = 10.1088/0957-4484/21/17/175202| pmid = 20368686| title = Nanotube devices based crossbar architecture: Toward neuromorphic computing| journal = Nanotechnology| volume = 21| issue = 17| pages = 175202| year = 2010| last1 = Zhao | first1 = W. S. | last2 = Agnus | first2 = G. | last3 = Derycke | first3 = V. | last4 = Filoramo | first4 = A. | last5 = Bourgoin | first5 = J. -P. | last6 = Gamrat | first6 = C. | bibcode = 2010Nanot..21q5202Z| url = https://zenodo.org/record/3428659}}</ref><ref name="humanbrainproject">{{YouTube|id=6RoiZ90mGfw|title=The Human Brain Project SP 9: Neuromorphic Computing Platform}}</ref> is the use of [[very-large-scale integration]] (VLSI) systems containing electronic [[analog circuit]]s to mimic neuro-biological architectures present in the nervous system. A neuromorphic computer/chip is any device that uses physical artificial neurons (made from silicon) to do computations.<ref name=":5">{{cite journal|last1=Mead|first1=Carver|title=Neuromorphic electronic systems|journal=Proceedings of the IEEE|date=1990|volume=78|issue=10|pages=1629–1636|doi=10.1109/5.58356|url=https://authors.library.caltech.edu/53090/1/00058356.pdf}}</ref><ref name=":2" /> In recent times, the term ''neuromorphic'' has been used to describe analog, digital, mixed-mode analog/digital VLSI, and software systems that implement models of [[neural system]]s (for [[perception]], [[motor control]], or [[multisensory integration]]). The implementation of neuromorphic computing on the hardware level can be realized by oxide-based [[memristor]]s,<ref name="Maan 1–13">{{Cite journal|last1=Maan|first1=A. K.|last2=Jayadevi|first2=D. A.|last3=James|first3=A. P.|date=2016-01-01|title=A Survey of Memristive Threshold Logic Circuits|journal=IEEE Transactions on Neural Networks and Learning Systems|volume=PP|issue=99|pages=1734–1746|doi=10.1109/TNNLS.2016.2547842|pmid=27164608|issn=2162-237X|arxiv=1604.07121|bibcode=2016arXiv160407121M|s2cid=1798273}}</ref> spintronic memories, threshold switches, and [[transistor]]s.<ref name=":6">{{Cite journal|title = Mott Memory and Neuromorphic Devices|journal = Proceedings of the IEEE|date = 2015-08-01|issn = 0018-9219|pages = 1289–1310|volume = 103|issue = 8|doi = 10.1109/JPROC.2015.2431914|first1 = You|last1 = Zhou|first2 = S.|last2 = Ramanathan|s2cid = 11347598|url = https://zenodo.org/record/895565}}</ref><ref name=":2">{{Cite document|title=Neuromorphic Circuits With Neural Modulation Enhancing the Information Content of Neural Signaling {{!}} International Conference on Neuromorphic Systems 2020|language=EN|doi=10.1145/3407197.3407204|s2cid=220794387}}</ref> Training software-based neuromorphic systems of spiking neural networks can be achieved using error backpropagation, e.g., using [[Python (programming language)|Python]] based frameworks such as snnTorch,<ref name=":7">{{cite journal |last1=Eshraghian|first1=Jason K.|last2=Ward|first2=Max|last3=Neftci |first3=Emre|last4=Wang|first4=Xinxin|last5=Lenz|first5=Gregor|last6=Dwivedi|first6=Girish|last7=Bennamoun|first7=Mohammed|last8=Jeong|first8=Doo Seok|last9=Lu|first9=Wei D.|title=Training Spiking Neural Networks Using Lessons from Deep Learning |date=1 October 2021 |arxiv=2109.12894 }}</ref> or using canonical learning rules from the biological learning literature, e.g., using BindsNet.<ref name=":8">{{Cite web | url=https://github.com/Hananel-Hazan/bindsnet | title=Hananel-Hazan/bindsnet: Simulation of spiking neural networks (SNNs) using PyTorch.| date=31 March 2020}}</ref><br />
<br />
<br />
'''<font color="#ff8000">神经形态工程Neuromorphic engineering</font>'''(也称为'''<font color="#ff8000">神经形态计算Neuromorphic computing</font>''')<ref name=":3" /><ref name=":4" /><ref name="humanbrainproject" />是指使用包含电子'''<font color="#ff8000">模拟电路Analog circuit</font>'''的'''<font color="#ff8000">超大规模集成电路Very-large-scale integration</font>'''系统来模拟神经系统中生理结构的研究方法。神经形态计算机或神经形态芯片包括任何使用由硅制成的人造神经元进行计算的设备。<ref name=":5" /><ref name=":2" />近年来，神经形态学（neuromorphic）这个术语被用来描述能够实现'''<font color="#ff8000">神经系统Neural system</font>'''模型功能(如'''<font color="#ff8000">感知Perception</font>'''、'''<font color="#ff8000">运动控制Motor control</font>'''，'''<font color="#ff8000">多感官整合Multisensory integration</font>'''等)的模拟、数字、模拟/数字混合模式超大规模集成电路和软件系统。神经形态计算的硬件实现可以通过基于氧化物的'''<font color="#ff8000">记忆电阻器Memristor</font>'''、<ref name="Maan 1–13" />自旋电子存储器、阈值开关和'''<font color="#ff8000">晶体管Transistor</font>'''来实现。<ref name=":6" /><ref name=":2" />对基于软件的脉冲神经网络系统的训练可以通过误差反向传播机制来实现，例如，使用snnTorch等基于Python的框架，<ref name=":7" />或使用BindsNet等典型的受生物启发的学习模式。<ref name=":8" /><br />
<br />
A key aspect of neuromorphic engineering is understanding how the morphology of individual neurons, circuits, applications, and overall architectures creates desirable computations, affects how information is represented, influences robustness to damage, incorporates learning and development, adapts to local change (plasticity), and facilitates evolutionary change.<br />
<br />
<br />
神经形态工程领域的一个关键问题，就是理解'''<font color="#32CD32">单个神经元形态、神经回路、应用和整体结构</font>'''如何产生理想的计算，如何影响信息的表示和对破坏的鲁棒性，如何整合学习和发展，如何适应局部变化(可塑性) 并促进逐渐发展的变化。<br />
<br />
Neuromorphic engineering is an interdisciplinary subject that takes inspiration from [[biology]], [[physics]], [[mathematics]], [[computer science]], and [[electronic engineering]]<ref name=":2" /> to design artificial neural systems, such as [[Machine vision|vision systems]], head-eye systems, auditory processors, and autonomous robots, whose physical architecture and design principles are based on those of biological nervous systems.<ref name=":9">{{Cite journal | doi = 10.1155/2012/705483| title = Qualitative Functional Decomposition Analysis of Evolved Neuromorphic Flight Controllers| journal = Applied Computational Intelligence and Soft Computing| volume = 2012| pages = 1–21| year = 2012| last1 = Boddhu | first1 = S. K. | last2 = Gallagher | first2 = J. C. | doi-access = free}}</ref> It was developed by [[Carver Mead]]<ref name=":10">{{cite web|title=carver mead website|last1=Mead|first1=Carver|url=http://www.carvermead.caltech.edu/index.html|website=carvermead}}</ref> in the late 1980s.<br />
<br />
<br />
神经形态工程是以'''<font color="#ff8000">生物学Biology</font>'''、'''<font color="#ff8000">物理学Physics</font>'''、'''<font color="#ff8000">数学Mathematics</font>'''、'''<font color="#ff8000">计算机科学Computer science</font>'''和'''<font color="#ff8000">电子工程Electronic engineering<ref name=":2" /></font>'''等学科为基础，设计人工神经系统（如'''<font color="#ff8000">视觉系统Vision systems</font>'''、头眼系统、听觉处理器以及物理结构和设计原则都受启发于生物神经系统的自主机器人）的一门交叉学科。<ref name=":9" />20世纪80年代后期，卡弗·米德极大地推动了神经形态工程领域的发展。<ref name=":10" /><br />
<br />
==Neurological inspiration==<br />
==神经生物学启发==<br />
Neuromorphic engineering is set apart by the inspiration it takes from what we know about the structure and operations of the [[brain]]. Neuromorphic engineering translates what we know about the brain's function into computer systems. Work has mostly focused on replicating the analog nature of [[biological computation]] and the role of [[neuron]]s in [[cognition]].<br />
<br />
<br />
神经形态工程形成的灵感来源于现已知的'''<font color="#ff8000">大脑Brain</font>'''结构及运作机制知识，它将我们对大脑功能的了解用于研究和优化计算机系统。该领域工作主要集中于对'''<font color="#ff8000">生物计算Biological computation</font>'''模拟特性和'''<font color="#ff8000">神经元Neuron</font>'''在'''<font color="#ff8000">认知Cognition</font>'''中所发挥作用的复刻。<br />
<br />
The biological processes of neurons and their [[synapse]]s are dauntingly complex, and thus very difficult to artificially simulate. A key feature of biological brains is that all of the processing in neurons use analog chemical signals. This makes it hard to replicate brains in computers because the current generation of computers is completely digital. However, the characteristics of these parts can be abstracted into mathematical functions that closely capture the essence of the neuron's operations.<br />
<br />
神经元及其'''<font color="#ff8000">突触Synapse</font>'''的生理学过程极其复杂，难以进行人工模拟。神经元在所有处理过程中都使用化学模拟信号，这是大脑的一个关键生理学特征。这个特征大大增加了在计算机中复制大脑的难度，因为目前的计算机是完全数字化的。然而，部分特征可以抽象为数学函数，这些函数能够紧密捕捉神经元运行的本质。<br />
<br />
The goal of neuromorphic computing is not to perfectly mimic the brain and all of its functions, but instead to extract what is known of its structure and operations to be used in a practical computing system. No neuromorphic system will claim nor attempt to reproduce every element of neurons and synapses, but all adhere to the idea that computation is highly [[distributed processing|distributed]] throughout a series of small computing elements analogous to a neuron. While this sentiment is standard, researchers chase this goal with different methods.<ref name=":11">{{Cite journal | doi = 10.1088/1741-2560/13/5/051001| title = Large-scale neuromorphic computing systems| journal = Journal of Neural Engineering| volume = 13| pages = 1–15| year = 2016| last1 = Furber | first1 = Steve| issue = 5| pmid = 27529195| bibcode = 2016JNEng..13e1001F| doi-access = free}}</ref><br />
<br />
神经形态计算的目标不是完美地模拟大脑及其所有功能，而是利用已知的大脑结构和运转机制来研发或优化实际的计算系统。任何神经形态学系统都不会声称或试图复制神经元和突触中的每一个元素，但所有人都一致认可将计算高度'''<font color="#ff8000">分散Distribute</font>'''于一系列类似于神经元的小型计算元素的理念。研究人员用不同的方法来追求这一普遍目标。<ref name=":11" /><br />
<br />
<br />
==Examples==<br />
==代表性成果==<br />
As early as 2006, researchers at [[Georgia Tech]] published a field programmable neural array.<ref name=":12">{{Cite book|title = A field programmable neural array|last1 = Farquhar|first1 = Ethan|date = May 2006|journal = IEEE International Symposium on Circuits and Systems|pages = 4114–4117|last2 = Hasler|first2 = Paul.|doi = 10.1109/ISCAS.2006.1693534|isbn = 978-0-7803-9389-9|s2cid = 206966013}}</ref> This chip was the first in a line of increasingly complex arrays of floating gate transistors that allowed programmability of charge on the gates of [[MOSFET]]s to model the channel-ion characteristics of neurons in the brain and was one of the first cases of a silicon programmable array of neurons.<br />
<br />
早在2006年，佐治亚理工学院的研究人员就研发出了一种现场可编程神经阵列。<ref name=":12" />在此之后，一系列越来越复杂的浮栅晶体管阵列被成功研发出来，这些晶体管阵列可以通过在'''<font color="#ff8000">金属-氧化物半导体效应晶体管MOSFET</font>'''的栅极上编程来模拟大脑中神经元的离子通道特性，这也是以硅为主要材料的可编程神经元阵列的最早成功案例之一。<br />
<br />
In November 2011, a group of [[MIT]] researchers created a computer chip that mimics the analog, ion-based communication in a synapse between two neurons using 400 transistors and standard [[CMOS]] manufacturing techniques.<ref name=":13">{{cite web|title=MIT creates "brain chip"|url=http://www.extremetech.com/extreme/105067-mit-creates-brain-chip|access-date=4 December 2012}}</ref><ref name="Neuromorphic silicon paper">{{cite journal|title=Neuromorphic silicon neurons and large-scale neural networks: challenges and opportunities|doi=10.3389/fnins.2011.00108|pmid=21991244|pmc=3181466|volume=5|pages=108|journal=Frontiers in Neuroscience|year=2011|last1=Poon|first1=Chi-Sang|last2=Zhou|first2=Kuan|doi-access=free}}</ref><br />
<br />
2011年11月，麻省理工学院的一组研究人员研发出一种计算机芯片，该芯片上使用标准的'''<font color="#ff8000">互补金属氧化物半导体CMOS</font>'''制造技术集成了400个晶体管，能够模拟神经元间突触中基于离子的通讯。<ref name=":13" /><ref name="Neuromorphic silicon paper" /><br />
<br />
In June 2012, [[spintronic]] researchers at [[Purdue University]] presented a paper on the design of a neuromorphic chip using [[Spin valve|lateral spin valve]]s and [[memristor]]s. They argue that the architecture works similarly to neurons and can therefore be used to test methods of reproducing the brain's processing. In addition, these chips are significantly more energy-efficient than conventional ones.<ref name="Spin Devices Prop">{{Cite arXiv|title=Proposal For Neuromorphic Hardware Using Spin Devices|eprint=1206.3227|last1=Sharad|first1=Mrigank|last2=Augustine|first2=Charles|last3=Panagopoulos|first3=Georgios|last4=Roy|first4=Kaushik|class=cond-mat.dis-nn|year=2012}}</ref><br />
<br />
2012年6月，普渡大学的'''<font color="#ff8000">自旋电子学Spintronic</font>'''研究人员发表了一篇关于利用'''<font color="#ff8000">侧向自旋阀Lateral spin valves</font>'''和'''<font color="#ff8000">记忆电阻器Memristors</font>'''设计神经形态芯片的论文。他们认为，这种芯片结构的工作原理与神经元相似，因此可以用于大脑运行机制的复刻方法的测试。此外，这些芯片在能耗方面明显优于传统芯片。<ref name="Spin Devices Prop" /><br />
<br />
Research at [[HP Labs]] on Mott memristors has shown that while they can be non-[[Volatile memory|volatile]], the volatile behavior exhibited at temperatures significantly below the [[phase transition]] temperature can be exploited to fabricate a [[neuristor]],<ref name=":0" /> a biologically-inspired device that mimics behavior found in neurons.<ref name=":0">{{Cite journal | doi = 10.1038/nmat3510| pmid = 23241533| title = A scalable neuristor built with Mott memristors| journal = Nature Materials| volume = 12| issue = 2| pages = 114–7| year = 2012| last1 = Pickett | first1 = M. D. | last2 = Medeiros-Ribeiro | first2 = G. | last3 = Williams | first3 = R. S. | bibcode = 2013NatMa..12..114P| s2cid = 16271627| url = https://semanticscholar.org/paper/b6ba6f496ace2b947f111059663e76bb60e9efeb}}</ref> In September 2013, they presented models and simulations that show how the spiking behavior of these neuristors can be used to form the components required for a [[Turing machine]].<ref name=":14">{{cite journal|doi=10.1088/0957-4484/24/38/384002|title=Phase transitions enable computational universality in neuristor-based cellular automata|author1=Matthew D Pickett|author2=R Stanley Williams|name-list-style=amp|date=September 2013|publisher=IOP Publishing Ltd|journal=Nanotechnology|volume=24|issue=38|pmid=23999059|bibcode=2013Nanot..24L4002P|at=384002}}</ref><br />
<br />
'''<font color="#ff8000">惠普实验室HP labs</font>'''在莫特记忆电阻器上的研究表明，尽管它们可以是非'''<font color="#ff8000">易失性Volatile</font>'''的，但是在'''<font color="#ff8000">相变Phase transition</font>'''温度以下时表现出的易失性行为可以被用来制造'''<font color="#ff8000">类神经元电阻器Neuristor</font>'''(一种生物学启发的模仿神经元行为的硬件)<ref name=":0" />。2013年9月，他们通过模型和仿真展示了这些类神经元电阻器的脉冲行为如何产生'''<font color="#ff8000">图灵机Turing machine</font>'''的所需元素。<ref name=":14" /><br />
<br />
[[Neurogrid]], built by ''Brains in Silicon'' at [[Stanford University]],<ref name=":15">{{cite journal|last1=Boahen|first1=Kwabena|title=Neurogrid: A Mixed-Analog-Digital Multichip System for Large-Scale Neural Simulations|journal=Proceedings of the IEEE|date=24 April 2014|volume=102|issue=5|pages=699–716|doi=10.1109/JPROC.2014.2313565|s2cid=17176371}}</ref> is an example of hardware designed using neuromorphic engineering principles. The circuit board is composed of 16 custom-designed chips, referred to as NeuroCores. Each NeuroCore's analog circuitry is designed to emulate neural elements for 65536 neurons, maximizing energy efficiency. The emulated neurons are connected using digital circuitry designed to maximize spiking throughput.<ref name=":16">{{cite journal|doi=10.1038/503022a|pmid = 24201264|title = Neuroelectronics: Smart connections|journal = Nature|volume = 503|issue = 7474|pages = 22–4|year = 2013|last1 = Waldrop|first1 = M. Mitchell|bibcode = 2013Natur.503...22W|doi-access = free}}</ref><ref name=":17">{{cite journal|doi=10.1109/JPROC.2014.2313565|title = Neurogrid: A Mixed-Analog-Digital Multichip System for Large-Scale Neural Simulations|journal = Proceedings of the IEEE|volume = 102|issue = 5|pages = 699–716|year = 2014|last1 = Benjamin|first1 = Ben Varkey|last2 = Peiran Gao|last3 = McQuinn|first3 = Emmett|last4 = Choudhary|first4 = Swadesh|last5 = Chandrasekaran|first5 = Anand R.|last6 = Bussat|first6 = Jean-Marie|last7 = Alvarez-Icaza|first7 = Rodrigo|last8 = Arthur|first8 = John V.|last9 = Merolla|first9 = Paul A.|last10 = Boahen|first10 = Kwabena|s2cid = 17176371}}</ref><br />
<br />
'''<font color="#ff8000">神经栅格Neurogrid<ref name=":15" /></font>'''是由斯坦福大学Brains in Silicon公司研发的、使用神经形态工程原理设计的硬件。该电路板由16个定制设计的芯片组成(NeuroCores)。在设计中，每个NeuroCore芯片的模拟电路对65536个神经元的神经元素进行模拟，以最大限度地提高能量效率。模拟出的神经元通过设计的数字电路连接，以最大化脉冲吞吐量。<ref name=":16" /><ref name=":17" /><br />
<br />
A research project with implications for neuromorphic engineering is the [[Human Brain Project]] that is attempting to simulate a complete human brain in a supercomputer using biological data. It is made up of a group of researchers in neuroscience, medicine, and computing.<ref name=":18">{{cite web|title=Involved Organizations|url=http://www.humanbrainproject.eu/partners.html|access-date=22 February 2013|url-status=dead|archive-url=https://web.archive.org/web/20130302142627/http://www.humanbrainproject.eu/partners.html|archive-date=2 March 2013}}</ref> [[Henry Markram]], the project's co-director, has stated that the project proposes to establish a foundation to explore and understand the brain and its diseases, and to use that knowledge to build new computing technologies. The three primary goals of the project are to better understand how the pieces of the brain fit and work together, to understand how to objectively diagnose and treat brain diseases, and to use the understanding of the human brain to develop neuromorphic computers. That the simulation of a complete human brain will require a supercomputer a thousand times more powerful than today's encourages the current focus on neuromorphic computers.<ref name=":19">{{cite web|title=Human Brain Project|url=http://www.humanbrainproject.eu|access-date=22 February 2013}}</ref> $1.3 billion has been allocated to the project by The [[European Commission]].<ref name=":20">{{cite web|title=The Human Brain Project and Recruiting More Cyberwarriors|url=http://www.marketplace.org/topics/tech/human-brain-project-and-recruiting-more-cyberwarriors|access-date=22 February 2013|date=January 29, 2013}}</ref><br />
<br />
'''<font color="#ff8000">人类大脑计划Human Brain Project</font>'''对神经形态工程具有较大影响，其主要任务是尝试用生物数据在超级计算机中模拟完整的人脑。人类大脑计划由神经科学、医学和计算机科学背景的研究人员组成。<ref name=":18" />该项目的联合主管亨利•马克拉姆(Henry Markram)表示，人类大脑计划的目的是建立一个探索和了解脑科学和脑疾病知识的基础，并利用这些知识来构建更先进的计算机技术。这个项目的三个主要目标分别是: 更好地理解大脑的各个部分是如何相互配合协同工作的; 理解如何客观地诊断和治疗脑部疾病; 以及利用对人类大脑的理解来开发神经形态计算机。模拟一个完整的人类大脑需要一台比现在强大一千倍的超级计算机，这不断激发着对神经形态计算机领域的研究兴趣。<ref name=":19" />欧盟委员会已经向人类大脑计划拨款13亿美元。<ref name=":20" /><br />
<br />
Other research with implications for neuromorphic engineering involves the [[BRAIN Initiative]]<ref name="economist">[https://www.economist.com/news/science-and-technology/21582495-computers-will-help-people-understand-brains-better-and-understanding-brains Neuromorphic computing: The machine of a new soul], The Economist, 2013-08-03</ref> and the [[TrueNorth]] chip from [[IBM]].<ref name=":21">{{cite journal|last1=Modha|first1=Dharmendra|title=A million spiking-neuron integrated circuit with a scalable communication network and interface|journal=Science|date=Aug 2014|volume=345|issue=6197|pages=668–673|doi=10.1126/science.1254642|pmid=25104385|bibcode=2014Sci...345..668M|s2cid=12706847}}</ref> Neuromorphic devices have also been demonstrated using nanocrystals, nanowires, and conducting polymers.<ref name=":22">{{Cite web|url=http://jessamynfairfield.com/wp-content/uploads/2017/03/PWMar17Fairfield.pdf|title=Smarter Machines|last=Fairfield|first=Jessamyn|date=March 1, 2017}}</ref><br />
<br />
其他与神经形态工程有关的研究还包括'''<font color="#ff8000">脑计划BRAIN initiative</font>'''，<ref name="economist" />和IBM研发的'''<font color="#ff8000">TrueNorth</font>'''芯片。<ref name=":21" />使用纳米晶体、纳米线和导电聚合物也能够用于制造神经形态学硬件。<ref name=":22" /><br />
<br />
[[Intel]] unveiled its neuromorphic research chip, called “[[Intel Loihi|Loihi]]”, in October 2017. The chip uses an asynchronous [[spiking neural network]] (SNN) to implement adaptive self-modifying event-driven fine-grained parallel computations used to implement learning and inference with high efficiency.<ref name=":23">{{cite journal |last1=Davies |first1=Mike |title=Loihi: A Neuromorphic Manycore Processor with On-Chip Learning |journal=IEEE Micro |date=January 16, 2018 |volume=38 |issue=1 |pages=82–99 |display-authors=etal|doi=10.1109/MM.2018.112130359 |s2cid=3608458 }}</ref><ref name="Morris2017">{{cite web |last1=Morris |first1=John |title=Why Intel built a neuromorphic chip |url=https://www.zdnet.com/article/why-intel-built-a-neuromorphic-chip/ |website=ZDNet |access-date=17 August 2018 |language=en}}</ref><br />
<br />
2017年10月，英特尔发布了神经形态芯片'''<font color="#ff8000">Loihi</font>'''。该芯片采用异步'''<font color="#ff8000">脉冲神经网络Spiking neural network</font>'''实现了自适应、自修改、事件驱动的细粒度并行计算，实现了高效的学习和推理。<ref name=":23" /><ref name="Morris2017" /><br />
<br />
[[IMEC]], a Belgium-based nanoelectronics research center, demonstrated the world's first self-learning neuromorphic chip. The brain-inspired chip, based on OxRAM technology, has the capability of self-learning and has been demonstrated to have the ability to compose music.<ref name=":24">{{cite web |title=Imec demonstrates self-learning neuromorphic chip that composes music |url=https://www.imec-int.com/en/articles/imec-demonstrates-self-learning-neuromorphic-chip-that-composes-music |website=IMEC International |access-date=1 October 2019}}</ref> IMEC released the 30-second tune composed by the prototype. The chip was sequentially loaded with songs in the same time signature and style. The songs were old Belgian and French flute minuets, from which the chip learned the rules at play and then applied them.<ref name=":25">{{cite web|last1=Bourzac|first1=Katherine|title=A Neuromorphic Chip That Makes Music|url=https://spectrum.ieee.org/a-neuromorphic-chip-that-makes-music|url-status=live|access-date=1 October 2019|website=IEEE Spectrum|date=May 23, 2017}}</ref><br />
<br />
比利时的'''<font color="#ff8000">微电子研究中心IMEC</font>'''研发了世界上首个自学习神经形态芯片。这种基于'''<font color="#32CD32"> OxRAM(filamentary-oxide-based resistive memory technology)</font>''' 技术的大脑启发芯片具有自学习能力，并且已被证明具有创作音乐的能力。<ref name=":24" />IMEC发布了由芯片原型机谱写的30秒旋律。向芯片加载一系列特征、风格统一的歌曲（古代比利时和法国长笛小步舞曲），芯片从中学习相关规则并将其应用于创作。<ref name=":25" /><br />
<br />
[[Blue Brain Project|The Blue Brain Project]], led by Henry Markram, aims to build biologically detailed digital reconstructions and simulations of the mouse brain. The Blue Brain Project has created in silico models of rodent brains, while attempting to replicate as many details about its biology as possible. The supercomputer-based simulations offer new perspectives on understanding the structure and functions of the brain.<br />
<br />
由亨利·马克拉姆领导的'''<font color="#ff8000">蓝脑计划The Blue Brain Project</font>'''旨在建立小鼠大脑生理学细节的数字重建和模拟。蓝脑计划已经建立了啮齿动物大脑的电子模型，同时进行着尽可能多地复制其生理学细节的尝试。基于超级计算机的模拟为理解大脑的结构和功能提供了新的视角。<br />
<br />
The European Union funded a series of projects at the University of Heidelberg, which led to the development of <br />
[[BrainScaleS]] (brain-inspired multiscale computation in neuromorphic hybrid systems), a hybrid analog [[neuromorphic]] supercomputer located at Heidelberg University, Germany. It was developed as part of the [[Human Brain Project]] neuromorphic computing platform and is the complement to the [[SpiNNaker]] supercomputer (which is based on digital technology). The architecture used in BrainScaleS mimics biological neurons and their connections on a physical level; additionally, since the components are made of silicon, these model neurons operate on average 864 times (24 hours of real time is 100 seconds in the machine simulation) that of their biological counterparts.<ref name=":26">{{Cite web|date=2016-03-21|title=Beyond von Neumann, Neuromorphic Computing Steadily Advances|url=https://www.hpcwire.com/2016/03/21/lacking-breakthrough-neuromorphic-computing-steadily-advance/|access-date=2021-10-08|website=HPCwire|language=en-US}}</ref><br />
<br />
欧盟资助了海德堡大学一系列促进BrainScaleS(神经形态混合系统中受大脑启发的多尺度计算)发展的项目，这是一台位于德国海德堡大学的混合模拟'''<font color="#ff8000">神经形态Neuromorphic</font>'''超级计算机。它是作为人类大脑计划中神经形态计算平台的一部分而开发的，是'''<font color="#ff8000">SpiNNaker</font>'''超级计算机(基于数字技术)的补充。BrainScaleS中使用的体系架构模拟了生物神经元及其在物理层面上的连接；此外，由于这些组件是由硅制成的，这些模型神经元平均运行速度是生物神经元的864倍，这意味着在机器模拟中，24小时的实时时间仅为100秒。<ref name=":26" /><br />
<br />
===Neuromorphic sensors===<br />
===神经形态传感器===<br />
The concept of neuromorphic systems can be extended to sensors (not just to computation). An example of this applied to detecting [[light]] is the [[retinomorphic sensor]] or, when employed in an array, the [[event camera]].<br />
<br />
神经形态系统的概念可以扩展到传感器(而不仅仅是计算单元)。一个用于检测光线的例子是'''<font color="#ff8000">类视网膜传感器Retinomorphic sensor</font>'''，或者'''<font color="#ff8000">事件摄像机Event camera</font>阵列。'''<br />
<br />
<br />
==Ethical considerations ==<br />
==伦理问题==<br />
While the interdisciplinary concept of neuromorphic engineering is relatively new, many of the same ethical considerations apply to neuromorphic systems as apply to [[human-like machines]] and [[artificial intelligence]] in general. However, the fact that neuromorphic systems are designed to mimic a [[human brain]] gives rise to unique ethical questions surrounding their usage.<br />
<br />
虽然神经形态工程这个跨学科概念相对较新，但许多适用于'''<font color="#ff8000">类人机器Human-like machines</font>'''和人工智能的伦理考虑大体上也适用于神经形态系统。另一方面，神经形态系统是为了模仿人类大脑而设计，这一事实引起了一些针对性的独特伦理问题。<br />
<br />
However, the practical debate is that neuromorphic hardware as well as artificial "neural networks" are immensely simplified models of how the brain operates or processes information at a much lower [[complex system|complexity in terms of size and functional technology]] and a much more regular structure in terms of [[brain connectivity|connectivity]]. Comparing [[neuromorphic chip]]s to the brain is a very crude comparison similar to comparing a plane to a bird just because they both have wings and a tail. The fact is that neural cognitive systems are many orders of magnitude more [[Energy efficiency (physics)|energy-]] and compute-efficient than current state-of-the-art AI and neuromorphic engineering is an attempt to narrow this gap by inspiring from the brain's mechanism just like many engineering designs have [[bioengineering|bio-inspired features]].<br />
<br />
然而，这种争论的实际情况是，神经形态硬件和人工“神经网络”是对大脑运作或处理信息过程的极其简化的模型，在大小和功能技术方面的'''<font color="#ff8000">复杂性Complexity</font>'''相比而言比较低，在'''<font color="#ff8000">连接性Connectivity</font>'''方面也具有更加规则的结构。将'''<font color="#ff8000">神经形态芯片Neuromorphic chips</font>'''与大脑进行比较是一种非常粗糙的比较，类似于仅仅因为飞机有翅膀和尾巴就将它与鸟进行比较。事实上，当前最先进的人工智能在'''<font color="#ff8000">能耗效率Energy-efficiency</font>'''和计算效率方面距离人脑神经认知系统仍有较大差距，而神经形态工程只是一种通过从大脑机制中获得灵感来缩小这种差距的尝试，就像许多工程设计中都具有'''<font color="#ff8000">生物启发的特征Bio-inspired features</font>'''一样。<br />
<br />
===Democratic concerns===<br />
===公众担忧===<br />
Significant ethical limitations may be placed on neuromorphic engineering due to public perception.<ref name=":27">{{Cite report|url=https://ai100.stanford.edu/sites/g/files/sbiybj9861/f/ai_100_report_0831fnl.pdf|title=Artificial Intelligence and Life in 2030|author=2015 Study Panel|date=September 2016|work=One Hundred Year Study on Artificial Intelligence (AI100)|publisher=Stanford University}}</ref> Special [[Eurobarometer]] 382: Public Attitudes Towards Robots, a survey conducted by the European Commission, found that 60% of [[European Union]] citizens wanted a ban of robots in the care of children, the elderly, or the disabled. Furthermore, 34% were in favor of a ban on robots in education, 27% in healthcare, and 20% in leisure. The European Commission classifies these areas as notably “human.” The report cites increased public concern with robots that are able to mimic or replicate human functions. Neuromorphic engineering, by definition, is designed to replicate the function of the human brain.<ref name=":1">{{Cite web|url=http://ec.europa.eu/commfrontoffice/publicopinion/archives/ebs/ebs_382_en.pdf|title=Special Eurobarometer 382: Public Attitudes Towards Robots|last=European Commission|date=September 2012|website=European Commission}}</ref><br />
<br />
由于公众认知的相关忧虑，神经形态工程学可能会受到严重的伦理限制。<ref name=":27" />欧盟委员会进行的一项调查发现，60% 的欧盟公民希望禁止机器人参与照顾儿童、老人或残疾人的工作。此外，34% 的人支持禁止机器人用于教育，27% 的人支持禁止机器人用于医疗保健，20% 的人支持禁止机器人用于休闲。欧盟委员会将以上领域划入“人类”范畴。报告指出，公众越来越关注能够模仿或复制人类功能的机器人。神经形态工程，顾名思义，是为了复制人脑的功能而设计的。<ref name=":1" /><br />
<br />
The democratic concerns surrounding neuromorphic engineering are likely to become even more profound in the future. The European Commission found that EU citizens between the ages of 15 and 24 are more likely to think of robots as human-like (as opposed to instrument-like) than EU citizens over the age of 55. When presented an image of a robot that had been defined as human-like, 75% of EU citizens aged 15–24 said it corresponded with the idea they had of robots while only 57% of EU citizens over the age of 55 responded the same way. The human-like nature of neuromorphic systems, therefore, could place them in the categories of robots many EU citizens would like to see banned in the future.<ref name=":1" /><br />
<br />
围绕神经形态工程的公众担忧可能在未来变得更加深刻。欧盟委员会发现，相比于55岁以上的欧盟公民，15至24岁的欧盟公民更有可能认为机器人像人(而不是像仪器)。当看到一张“类人”机器人的图片时，年龄在15岁至24岁之间的欧盟公民中有75% 的人表示这符合他们对机器人的想法，而55岁以上的欧盟公民中只有57% 的人有同样的反应。因此，神经形态系统可能因为其类似人类的特性而被归入许多欧盟公民希望在未来禁止使用的机器人类别。<ref name=":1" /><br />
<br />
=== Personhood===<br />
=== 人格问题===<br />
As neuromorphic systems have become increasingly advanced, some scholars{{who|date=August 2021}} have advocated for granting [[personhood]] rights to these systems. If the brain is what grants humans their personhood, to what extent does a neuromorphic system have to mimic the human brain to be granted personhood rights? Critics of technology development in the [[Human Brain Project]], which aims to advance brain-inspired computing, have argued that advancement in neuromorphic computing could lead to machine consciousness or personhood.<ref name=":28">{{Cite journal|last=Aicardi|first=Christine|date=September 2018|title=Accompanying technology development in the Human Brain Project: From foresight to ethics management|journal=Futures|volume=102|pages=114–124|doi=10.1016/j.futures.2018.01.005|doi-access=free}}</ref> If these systems are to be treated as people, critics argue, then many tasks humans perform using neuromorphic systems, including the act of termination of neuromorphic systems, may be morally impermissible as these acts would violate the autonomy of the neuromorphic systems.<ref name=":29">{{Cite journal|last=Lim|first=Daniel|date=2014-06-01|title=Brain simulation and personhood: a concern with the Human Brain Project|journal=Ethics and Information Technology|language=en|volume=16|issue=2|pages=77–89|doi=10.1007/s10676-013-9330-5|s2cid=17415814|issn=1572-8439}}</ref><br />
<br />
随着神经形态系统的日益发展，一些学者主张赋予这些系统'''<font color="ff8000">人格Personhood</font>'''权。如果是大脑赋予了人类人格，那么在多大程度上模仿人类大脑的神经形态系统才能被赋予人格权利？“人类大脑计划”旨在推进以大脑为灵感的计算机技术发展，该计划的批评者认为，神经形态计算机技术的进步可能导致机器意识或人格的形成。<ref name=":28" />这些批评者认为，如果这些系统被当作人来对待，那么人类使用神经形态系统执行任务（包括终止神经形态系统）的行为，在道德上就可能是不被允许的，因为这些行为将违反神经形态系统的自主性。<ref name=":29" /><br />
<br />
<br />
==Dual use (military applications)==<br />
==军民两用技术（军事应用）==<br />
The [[Joint Artificial Intelligence Center]], a branch of the U.S. military, is a center dedicated to the procurement and implementation of AI software and neuromorphic hardware for combat use. Specific applications include smart headsets/goggles and robots. JAIC intends to rely heavily on neuromorphic technology to connect "every fighter every shooter" within a network of neuromorphic-enabled units.<br />
<br />
'''<font color="#ff8000">联合人工智能中心The Joint Artificial Intelligence Center</font>'''(JAIC),是美国军队的一个分支，专门从事采购和实施用于战斗的人工智能软件和神经形态硬件。具体应用包括智能耳机/护目镜和机器人。JAIC打算高度依赖神经形态技术，'''<font color="#32CD32">使用神经形态技术来连接神经形态单位网络中的“每个战士每个射手”</font>'''。<br />
<br />
<br />
==Legal considerations==<br />
==法律问题==<br />
Skeptics have argued that there is no way to apply the electronic personhood, the concept of personhood that would apply to neuromorphic technology, legally. In a letter signed by 285 experts in law, robotics, medicine, and ethics opposing a European Commission proposal to recognize “smart robots” as legal persons, the authors write, “A legal status for a robot can’t derive from the [[Natural person|Natural Person]] model, since the robot would then hold [[human rights]], such as the right to dignity, the right to its integrity, the right to remuneration or the right to citizenship, thus directly confronting the Human rights. This would be in contradiction with the [[Charter of Fundamental Rights of the European Union]] and the [[Convention for the Protection of Human Rights and Fundamental Freedoms]].”<ref name=":30">{{Cite web|url=http://www.robotics-openletter.eu/|title=Robotics Openletter {{!}} Open letter to the European Commission|language=fr-FR|access-date=2019-05-10}}</ref><br />
<br />
质疑者者认为，在法律上没有办法应用能够适用于神经形态技术的电子人格。在一封由285名法律、机器人技术、医学和伦理学专家的联名信中，作者们反对欧盟委员会提出的承认“智能机器人”为法人的提议。他们写道，“机器人的法律地位无法从'''<font color="#ff8000">自然人Natural Person</font>'''模型中推导出来，因为机器人将被赋予'''<font color="#ff8000">人权Human rights</font>'''，如尊严权、完整权、报酬权或公民权，从而直接面临人权问题。这将有悖于《欧联基本权利宪章》和《欧洲保障人权和根本自由公约》”。<ref name=":30" /><br />
<br />
===Ownership and property rights ===<br />
===所有权及财产权问题===<br />
There is significant legal debate around property rights and artificial intelligence. In ''Acohs Pty Ltd v. Ucorp Pty Ltd'', Justice Christopher Jessup of the [[Federal Court of Australia]] found that the [[source code]] for [[Material safety data sheets|Material Safety Data Sheets]] could not be [[Copyright law of Australia|copyrighted]] as it was generated by a [[software interface]] rather than a human author.<ref name=":31">{{Cite web|url=http://www.lavan.com.au/advice/intellectual_property_technology/copyright_in_source_code_and_digital_products|title=Copyright in source code and digital products|last=Lavan|website=Lavan|language=en|access-date=2019-05-10}}</ref> The same question may apply to neuromorphic systems: if a neuromorphic system successfully mimics a human brain and produces a piece of original work, who, if anyone, should be able to claim ownership of the work?<ref name=":32">{{cite journal |last1=Eshraghian|first1=Jason K. |title=Human Ownership of Artificial Creativity |journal=Nature Machine Intelligence |date=9 March 2020 |volume=2 |pages=157–160 |doi=10.1038/s42256-020-0161-x}}</ref><br />
<br />
法律界围绕财产权和人工智能有着重大争论。在Acohs Pty有限公司诉Ucorp Pty有限公司一案中，澳大利亚联邦法院的克里斯托弗·杰瑟普法官发现，版权保护不适用于材料安全数据表的'''<font color="#ff8000">源代码Source code</font>'''，因为它是由'''<font color="#ff8000">软件界面Software interface</font>'''生成而非人类工作者生成的。<ref name=":31" />同样的问题可能也适用于神经形态系统: 如果一个神经形态系统成功地模仿了人类的大脑，并产生了一部原创作品，那么该如何确认这部作品的所有权归属？<ref name=":32" /><br />
<br />
<br />
==Neuromemristive systems==<br />
==神经记忆电阻系统==<br />
Neuromemristive systems are a subclass of neuromorphic computing systems that focus on the use of [[memristors]] to implement [[neuroplasticity]]. While neuromorphic engineering focuses on mimicking biological behavior, neuromemristive systems focus on abstraction.<ref name=":33">{{Cite web|url=https://digitalops.sandia.gov/Mediasite/Play/a10cf6ceb55d47608bb8326dd00e46611d|title=002.08 N.I.C.E. Workshop 2014: Towards Intelligent Computing with Neuromemristive Circuits and Systems - Feb. 2014|website=digitalops.sandia.gov|access-date=2019-08-26}}</ref> For example, a neuromemristive system may replace the details of a [[Cerebral cortex|cortical]] microcircuit's behavior with an abstract neural network model.<ref name=":34">C. Merkel and D. Kudithipudi, "Neuromemristive extreme learning machines for pattern classification," ISVLSI, 2014.</ref><br />
<br />
神经记忆电阻系统是神经形态计算系统的一个亚类，主要研究利用'''<font color="ff8000">记忆电阻器Memristors</font>'''实现'''<font color="ff8000">神经可塑性Neuroplasticity</font>'''。神经形态工程的重点是模拟生物行为，而神经记忆电阻系统的重点是提取。<ref name=":33" />举个例子，一个神经记忆系统可能用抽象的神经网络模型替代'''<font color="ff8000">皮层Cortical</font>'''微电路的行为细节。<ref name=":34" /><br />
<br />
There exist several neuron inspired threshold logic functions<ref name="Maan 1–13" /> implemented with memristors that have applications in high level [[pattern recognition]] applications. Some of the applications reported recently include [[speech recognition]],<ref name=":35">{{Cite journal|title = Memristor pattern recogniser: isolated speech word recognition|journal = Electronics Letters|pages = 1370–1372|volume = 51|issue = 17|doi = 10.1049/el.2015.1428|first1 = A.K.|last1 = Maan|first2 = A.P.|last2 = James|first3 = S.|last3 = Dimitrijev|year = 2015|bibcode = 2015ElL....51.1370M|hdl = 10072/140989|s2cid = 61454815|url = https://semanticscholar.org/paper/48d3ab11ec6e213b62f11eedcfb7b7febb058674|hdl-access = free}}</ref> [[face recognition]]<ref name=":36">{{Cite journal|title = Memristive Threshold Logic Face Recognition|journal = Procedia Computer Science|date = 2014-01-01|pages = 98–103|volume = 41|series = 5th Annual International Conference on Biologically Inspired Cognitive Architectures, 2014 BICA|doi = 10.1016/j.procs.2014.11.090|first1 = Akshay Kumar|last1 = Maan|first2 = Dinesh S.|last2 = Kumar|first3 = Alex Pappachen|last3 = James|doi-access = free}}</ref> and [[object recognition]].<ref name=":37">{{Cite journal|title = Memristive Threshold Logic Circuit Design of Fast Moving Object Detection|journal = IEEE Transactions on Very Large Scale Integration (VLSI) Systems|date = 2015-10-01|issn = 1063-8210|pages = 2337–2341|volume = 23|issue = 10|doi = 10.1109/TVLSI.2014.2359801|first1 = A.K.|last1 = Maan|first2 = D.S.|last2 = Kumar|first3 = S.|last3 = Sugathan|first4 = A.P.|last4 = James|arxiv = 1410.1267|s2cid = 9647290}}</ref> They also find applications in replacing conventional digital logic gates.<ref name=":38">{{Cite journal|title = Resistive Threshold Logic|journal = IEEE Transactions on Very Large Scale Integration (VLSI) Systems|date = 2014-01-01|issn = 1063-8210|pages = 190–195|volume = 22|issue = 1|doi = 10.1109/TVLSI.2012.2232946|first1 = A.P.|last1 = James|first2 = L.R.V.J.|last2 = Francis|first3 = D.S.|last3 = Kumar|arxiv = 1308.0090|s2cid = 7357110}}</ref><ref name=":39">{{Cite journal|title = Threshold Logic Computing: Memristive-CMOS Circuits for Fast Fourier Transform and Vedic Multiplication|journal = IEEE Transactions on Very Large Scale Integration (VLSI) Systems|date = 2015-11-01|issn = 1063-8210|pages = 2690–2694|volume = 23|issue = 11|doi = 10.1109/TVLSI.2014.2371857|first1 = A.P.|last1 = James|first2 = D.S.|last2 = Kumar|first3 = A.|last3 = Ajayan|arxiv = 1411.5255|s2cid = 6076956}}</ref><br />
<br />
受神经元启发、使用记忆电阻器实现的阈值逻辑函数<ref name="Maan 1–13" />在高级'''<font color="#ff8000">模式识别Pattern recognition</font>'''中有着广泛的应用，最近报道中其应用包括'''<font color="#ff8000">语音识别Speech recognition<ref name=":35" /></font>'''、'''<font color="#ff8000">人脸识别Face recognition<ref name=":36" /></font>'''和'''<font color="#ff8000">物体识别Object recognition<ref name=":37" /></font>'''。阈值逻辑函数还可以用来取代传统的数字逻辑门。<ref name=":38" /><ref name=":39" /><br />
<br />
For ideal passive memristive circuits there is an exact equation (Caravelli-Traversa-[[Di Ventra]] equation) for the internal memory of the circuit:<ref name=":40">{{cite journal |last=Caravelli |display-authors=etal|arxiv=1608.08651 |title=The complex dynamics of memristive circuits: analytical results and universal slow relaxation |year=2017 |doi=10.1103/PhysRevE.95.022140 |pmid= 28297937 |volume=95 |issue= 2 |pages= 022140 |journal=Physical Review E|bibcode=2017PhRvE..95b2140C |s2cid=6758362}}</ref><br />
<br />
对于理想的无源记忆电路，电路的内部记忆可以用精确的方程(Caravelli-Traversa-Di Ventra方程) 来描述:<ref name=":40" /><br />
<br />
:<math> \frac{d}{dt} \vec{W} = \alpha \vec{W}-\frac{1}{\beta} (I+\xi \Omega W)^{-1} \Omega \vec S </math><br />
<br />
:\frac{d}{dt} \vec{W} = \alpha \vec{W}-\frac{1}{\beta} (I+\xi \Omega W)^{-1} \Omega \vec S<br />
<br />
as a function of the properties of the physical memristive network and the external sources. In the equation above, <math>\alpha</math> is the "forgetting" time scale constant, <math> \xi=r-1</math> and <math>r=\frac{R_\text{off}}{R_\text{on}}</math> is the ratio of ''off'' and ''on'' values of the limit resistances of the memristors, <math> \vec S </math> is the vector of the sources of the circuit and <math>\Omega</math> is a projector on the fundamental loops of the circuit. The constant <math>\beta</math> has the dimension of a voltage and is associated to the properties of the [[memristor]]; its physical origin is the charge mobility in the conductor. The diagonal matrix and vector <math>W=\operatorname{diag}(\vec W)</math> and <math>\vec W</math> respectively, are instead the internal value of the memristors, with values between 0 and 1. This equation thus requires adding extra constraints on the memory values in order to be reliable.<br />
<br />
Caravelli-Traversa-Di Ventra方程是描述物理记忆网络和外部源性质的函数。在上述方程中，<math>\alpha</math>是“遗忘”时间尺度常数，<math>\xi=r-1</math>，<math>r =\frac{R\text_{off}}{R_\text{on}}</math>是记忆电阻器off状态和on状态极限电阻值之比，<math>\vec S</math>是电路源的矢量，<math>\Omega</math>是电路基本环路的投影。常数<math>\beta</math>具有电压的量纲，与记忆电阻器的特性有关；其物理原型是导体中的电荷迁移率。对角矩阵和向量 <math>W=\operatorname{diag}(\vec W)</math>和<nowiki><math>\vec W<math></nowiki>'''<font color="#32CD32">是记忆电阻器的内阻</font>'''，值在0到1之间。因此，这个等式需要在'''<font color="32CD32">内存值</font>'''上添加额外约束以保证可靠性。<br />
<br />
<br />
==See also==<br />
==相关词条==<br />
{{Columns-list|colwidth=18em|<br />
* [[AI accelerator (computer hardware)|AI accelerator]]<br />
* [[Artificial brain]]<br />
* [[Biomorphic]]<br />
* [[Cognitive computer]]<br />
* [[Computation and Neural Systems]]<br />
* [[Differentiable programming]]<br />
* [[Event camera]]<br />
* [[Neurorobotics]]<br />
* [[Optical flow sensor]]<br />
* [[Physical neural network]]<br />
* [[SpiNNaker]]<br />
* [[SyNAPSE]]<br />
* [[Retinomorphic sensor]]<br />
* [[Vision chip]]<br />
* [[Vision processing unit]]<br />
* [[Zeroth (software)]]<br />
* [[Hardware for artificial intelligence]]<br />
}}<br />
<br />
{{Portal bar|Electronics}}<br />
<br />
==References==<br />
{{Reflist|40em}}<br />
<br />
== External links==<br />
<!--======================== {{No more links}} ============================<br />
| PLEASE BE CAUTIOUS IN ADDING MORE LINKS TO THIS ARTICLE. Wikipedia |<br />
| is not a collection of links nor should it be used for advertising. |<br />
| |<br />
| Excessive or inappropriate links WILL BE DELETED. |<br />
| See [[Wikipedia:External links]] & [[Wikipedia:Spam]] for details. |<br />
| |<br />
| If there are already plentiful links, please propose additions or |<br />
| replacements on this article's discussion page, or submit your link |<br />
| to the relevant category at the Open Directory Project (dmoz.org) |<br />
| and link back to that category using the {{dmoz}} template. |<br />
======================={{No more links}}=============================--><br />
{{Commons category|Neuromorphic Engineering}}<br />
*[https://web.archive.org/web/20150727034331/http://ine-web.org/workshops/workshops-overview Telluride Neuromorphic Engineering Workshop]<br />
*[https://archive.today/20130115190057/http://capocaccia.ethz.ch/ CapoCaccia Cognitive Neuromorphic Engineering Workshop]<br />
*[https://web.archive.org/web/20190716132350/http://www.ine-web.org/ Institute of Neuromorphic Engineering]<br />
*[http://www.ine-news.org/ INE news site].<br />
*[http://www.frontiersin.org/neuromorphic_engineering Frontiers in Neuromorphic Engineering Journal]<br />
*[http://www.cns.caltech.edu/ Computation and Neural Systems] department at the [[California Institute of Technology]].<br />
*[http://www.humanbrainproject.eu/ Human Brain Project official site]<br />
*[https://www.the-scientist.com/features/building-a-silicon-brain-65738 Building a Silicon Brain:] Computer chips based on biological neurons may help simulate larger and more-complex brain models. May 1, 2019. SANDEEP RAVINDRAN<br />
<br />
<br />
*Telluride Neuromorphic Engineering Workshop<br />
*CapoCaccia Cognitive Neuromorphic Engineering Workshop<br />
* Institute of Neuromorphic Engineering<br />
* INE news site.<br />
* Frontiers in Neuromorphic Engineering Journal<br />
*Computation and Neural Systems department at the California Institute of Technology.<br />
*Human Brain Project official site<br />
*Building a Silicon Brain: Computer chips based on biological neurons may help simulate larger and more-complex brain models. May 1, 2019. SANDEEP RAVINDRAN<br />
<br />
=<nowiki>外部链接</nowiki>= <br />
*碲化物神经形态工程工作室<br />
*CapoCaccia认知神经形态工程工作室<br />
*神经形态工程研究所<br />
*INE新闻站点。<br />
*《神经形态工程学前沿》<br />
*加州理工学院计算与神经系统系。<br />
*人脑项目官方网站<br />
*打造硅脑: 基于生物神经元的计算机芯片可能有助于模拟更大、更复杂的大脑模型。2019年5月1日。SANDEEP RAVINDRAN.<br />
<br />
{{Differentiable computing}}<br />
{{Authority control}}<br />
<br />
[[Category:Electrical engineering]]<br />
[[Category:Neuroscience]]<br />
[[分类:AI accelerators|*]]<br />
[[Category:Artificial intelligence]]<br />
[[Category:Robotics]]<br />
<br />
Category:Electrical engineering<br />
Category:Neuroscience<br />
*<br />
Category:Artificial intelligence<br />
Category:Robotics<br />
<br />
类别: 电气工程类别: 神经科学 * 类别: 人工智能类别: 机器人<br />
<br />
<noinclude><br />
<br />
<small>This page was moved from [[wikipedia:en:Neuromorphic engineering]]. Its edit history can be viewed at [[神经形态计算/edithistory]]</small></noinclude><br />
<br />
[[Category:待整理页面]]<br />
<br />
<references /></div>Shenky20https://wiki.swarma.org/index.php?title=%E7%A5%9E%E7%BB%8F%E5%BD%A2%E6%80%81%E8%AE%A1%E7%AE%97&diff=30749神经形态计算2022-04-28T11:06:00Z<p>Shenky20：/* 法律问题 */</p>
<hr />
<div><br />
此词条由神经动力学模型读书会词条梳理志愿者Shenky20翻译审校，未经专家审校，带来阅读不便，请见谅。<br />
<br />
{{Use American English|date = January 2019}}<br />
{{Use mdy dates|date = January 2019}}<br />
<br />
'''Neuromorphic engineering''', also known as '''neuromorphic computing''',<ref name=":3">{{Cite journal | doi = 10.1145/2601069| title = Neuromorphic computing gets ready for the (really) big time| journal = [[Communications of the ACM]]| volume = 57| issue = 6| year = 2014| last1 = Monroe | first1 = D. | pages = 13–15| s2cid = 20051102}}</ref><ref name=":4">{{Cite journal | doi = 10.1088/0957-4484/21/17/175202| pmid = 20368686| title = Nanotube devices based crossbar architecture: Toward neuromorphic computing| journal = Nanotechnology| volume = 21| issue = 17| pages = 175202| year = 2010| last1 = Zhao | first1 = W. S. | last2 = Agnus | first2 = G. | last3 = Derycke | first3 = V. | last4 = Filoramo | first4 = A. | last5 = Bourgoin | first5 = J. -P. | last6 = Gamrat | first6 = C. | bibcode = 2010Nanot..21q5202Z| url = https://zenodo.org/record/3428659}}</ref><ref name="humanbrainproject">{{YouTube|id=6RoiZ90mGfw|title=The Human Brain Project SP 9: Neuromorphic Computing Platform}}</ref> is the use of [[very-large-scale integration]] (VLSI) systems containing electronic [[analog circuit]]s to mimic neuro-biological architectures present in the nervous system. A neuromorphic computer/chip is any device that uses physical artificial neurons (made from silicon) to do computations.<ref name=":5">{{cite journal|last1=Mead|first1=Carver|title=Neuromorphic electronic systems|journal=Proceedings of the IEEE|date=1990|volume=78|issue=10|pages=1629–1636|doi=10.1109/5.58356|url=https://authors.library.caltech.edu/53090/1/00058356.pdf}}</ref><ref name=":2" /> In recent times, the term ''neuromorphic'' has been used to describe analog, digital, mixed-mode analog/digital VLSI, and software systems that implement models of [[neural system]]s (for [[perception]], [[motor control]], or [[multisensory integration]]). The implementation of neuromorphic computing on the hardware level can be realized by oxide-based [[memristor]]s,<ref name="Maan 1–13">{{Cite journal|last1=Maan|first1=A. K.|last2=Jayadevi|first2=D. A.|last3=James|first3=A. P.|date=2016-01-01|title=A Survey of Memristive Threshold Logic Circuits|journal=IEEE Transactions on Neural Networks and Learning Systems|volume=PP|issue=99|pages=1734–1746|doi=10.1109/TNNLS.2016.2547842|pmid=27164608|issn=2162-237X|arxiv=1604.07121|bibcode=2016arXiv160407121M|s2cid=1798273}}</ref> spintronic memories, threshold switches, and [[transistor]]s.<ref name=":6">{{Cite journal|title = Mott Memory and Neuromorphic Devices|journal = Proceedings of the IEEE|date = 2015-08-01|issn = 0018-9219|pages = 1289–1310|volume = 103|issue = 8|doi = 10.1109/JPROC.2015.2431914|first1 = You|last1 = Zhou|first2 = S.|last2 = Ramanathan|s2cid = 11347598|url = https://zenodo.org/record/895565}}</ref><ref name=":2">{{Cite document|title=Neuromorphic Circuits With Neural Modulation Enhancing the Information Content of Neural Signaling {{!}} International Conference on Neuromorphic Systems 2020|language=EN|doi=10.1145/3407197.3407204|s2cid=220794387}}</ref> Training software-based neuromorphic systems of spiking neural networks can be achieved using error backpropagation, e.g., using [[Python (programming language)|Python]] based frameworks such as snnTorch,<ref name=":7">{{cite journal |last1=Eshraghian|first1=Jason K.|last2=Ward|first2=Max|last3=Neftci |first3=Emre|last4=Wang|first4=Xinxin|last5=Lenz|first5=Gregor|last6=Dwivedi|first6=Girish|last7=Bennamoun|first7=Mohammed|last8=Jeong|first8=Doo Seok|last9=Lu|first9=Wei D.|title=Training Spiking Neural Networks Using Lessons from Deep Learning |date=1 October 2021 |arxiv=2109.12894 }}</ref> or using canonical learning rules from the biological learning literature, e.g., using BindsNet.<ref name=":8">{{Cite web | url=https://github.com/Hananel-Hazan/bindsnet | title=Hananel-Hazan/bindsnet: Simulation of spiking neural networks (SNNs) using PyTorch.| date=31 March 2020}}</ref><br />
<br />
<br />
'''<font color="#ff8000">神经形态工程Neuromorphic engineering</font>'''(也称为'''<font color="#ff8000">神经形态计算Neuromorphic computing</font>''')<ref name=":3" /><ref name=":4" /><ref name="humanbrainproject" />是指使用包含电子'''<font color="#ff8000">模拟电路Analog circuit</font>'''的'''<font color="#ff8000">超大规模集成电路Very-large-scale integration</font>'''系统来模拟神经系统中生理结构的研究方法。神经形态计算机或神经形态芯片包括任何使用由硅制成的人造神经元进行计算的设备。<ref name=":5" /><ref name=":2" />近年来，神经形态学（neuromorphic）这个术语被用来描述能够实现'''<font color="#ff8000">神经系统Neural system</font>'''模型功能(如'''<font color="#ff8000">感知Perception</font>'''、'''<font color="#ff8000">运动控制Motor control</font>'''，'''<font color="#ff8000">多感官整合Multisensory integration</font>'''等)的模拟、数字、模拟/数字混合模式超大规模集成电路和软件系统。神经形态计算的硬件实现可以通过基于氧化物的'''<font color="#ff8000">记忆电阻器Memristor</font>'''、<ref name="Maan 1–13" />自旋电子存储器、阈值开关和'''<font color="#ff8000">晶体管Transistor</font>'''来实现。<ref name=":6" /><ref name=":2" />对基于软件的脉冲神经网络系统的训练可以通过误差反向传播机制来实现，例如，使用snnTorch等基于Python的框架，<ref name=":7" />或使用BindsNet等典型的受生物启发的学习模式。<ref name=":8" /><br />
<br />
A key aspect of neuromorphic engineering is understanding how the morphology of individual neurons, circuits, applications, and overall architectures creates desirable computations, affects how information is represented, influences robustness to damage, incorporates learning and development, adapts to local change (plasticity), and facilitates evolutionary change.<br />
<br />
<br />
神经形态工程领域的一个关键问题，就是理解'''<font color="#32CD32">单个神经元形态、神经回路、应用和整体结构</font>'''如何产生理想的计算，如何影响信息的表示和对破坏的鲁棒性，如何整合学习和发展，如何适应局部变化(可塑性) 并促进逐渐发展的变化。<br />
<br />
Neuromorphic engineering is an interdisciplinary subject that takes inspiration from [[biology]], [[physics]], [[mathematics]], [[computer science]], and [[electronic engineering]]<ref name=":2" /> to design artificial neural systems, such as [[Machine vision|vision systems]], head-eye systems, auditory processors, and autonomous robots, whose physical architecture and design principles are based on those of biological nervous systems.<ref name=":9">{{Cite journal | doi = 10.1155/2012/705483| title = Qualitative Functional Decomposition Analysis of Evolved Neuromorphic Flight Controllers| journal = Applied Computational Intelligence and Soft Computing| volume = 2012| pages = 1–21| year = 2012| last1 = Boddhu | first1 = S. K. | last2 = Gallagher | first2 = J. C. | doi-access = free}}</ref> It was developed by [[Carver Mead]]<ref name=":10">{{cite web|title=carver mead website|last1=Mead|first1=Carver|url=http://www.carvermead.caltech.edu/index.html|website=carvermead}}</ref> in the late 1980s.<br />
<br />
<br />
神经形态工程是以'''<font color="#ff8000">生物学Biology</font>'''、'''<font color="#ff8000">物理学Physics</font>'''、'''<font color="#ff8000">数学Mathematics</font>'''、'''<font color="#ff8000">计算机科学Computer science</font>'''和'''<font color="#ff8000">电子工程Electronic engineering<ref name=":2" /></font>'''等学科为基础，设计人工神经系统（如'''<font color="#ff8000">视觉系统Vision systems</font>'''、头眼系统、听觉处理器以及物理结构和设计原则都受启发于生物神经系统的自主机器人）的一门交叉学科。<ref name=":9" />20世纪80年代后期，卡弗·米德极大地推动了神经形态工程领域的发展。<ref name=":10" /><br />
<br />
==Neurological inspiration==<br />
==神经生物学启发==<br />
Neuromorphic engineering is set apart by the inspiration it takes from what we know about the structure and operations of the [[brain]]. Neuromorphic engineering translates what we know about the brain's function into computer systems. Work has mostly focused on replicating the analog nature of [[biological computation]] and the role of [[neuron]]s in [[cognition]].<br />
<br />
<br />
神经形态工程形成的灵感来源于现已知的'''<font color="#ff8000">大脑Brain</font>'''结构及运作机制知识，它将我们对大脑功能的了解用于研究和优化计算机系统。该领域工作主要集中于对'''<font color="#ff8000">生物计算Biological computation</font>'''模拟特性和'''<font color="#ff8000">神经元Neuron</font>'''在'''<font color="#ff8000">认知Cognition</font>'''中所发挥作用的复刻。<br />
<br />
The biological processes of neurons and their [[synapse]]s are dauntingly complex, and thus very difficult to artificially simulate. A key feature of biological brains is that all of the processing in neurons use analog chemical signals. This makes it hard to replicate brains in computers because the current generation of computers is completely digital. However, the characteristics of these parts can be abstracted into mathematical functions that closely capture the essence of the neuron's operations.<br />
<br />
神经元及其'''<font color="#ff8000">突触Synapse</font>'''的生理学过程极其复杂，难以进行人工模拟。神经元在所有处理过程中都使用化学模拟信号，这是大脑的一个关键生理学特征。这个特征大大增加了在计算机中复制大脑的难度，因为目前的计算机是完全数字化的。然而，部分特征可以抽象为数学函数，这些函数能够紧密捕捉神经元运行的本质。<br />
<br />
The goal of neuromorphic computing is not to perfectly mimic the brain and all of its functions, but instead to extract what is known of its structure and operations to be used in a practical computing system. No neuromorphic system will claim nor attempt to reproduce every element of neurons and synapses, but all adhere to the idea that computation is highly [[distributed processing|distributed]] throughout a series of small computing elements analogous to a neuron. While this sentiment is standard, researchers chase this goal with different methods.<ref name=":11">{{Cite journal | doi = 10.1088/1741-2560/13/5/051001| title = Large-scale neuromorphic computing systems| journal = Journal of Neural Engineering| volume = 13| pages = 1–15| year = 2016| last1 = Furber | first1 = Steve| issue = 5| pmid = 27529195| bibcode = 2016JNEng..13e1001F| doi-access = free}}</ref><br />
<br />
神经形态计算的目标不是完美地模拟大脑及其所有功能，而是利用已知的大脑结构和运转机制来研发或优化实际的计算系统。任何神经形态学系统都不会声称或试图复制神经元和突触中的每一个元素，但所有人都一致认可将计算高度'''<font color="#ff8000">分散Distribute</font>'''于一系列类似于神经元的小型计算元素的理念。研究人员用不同的方法来追求这一普遍目标。<ref name=":11" /><br />
<br />
<br />
==Examples==<br />
==代表性成果==<br />
As early as 2006, researchers at [[Georgia Tech]] published a field programmable neural array.<ref name=":12">{{Cite book|title = A field programmable neural array|last1 = Farquhar|first1 = Ethan|date = May 2006|journal = IEEE International Symposium on Circuits and Systems|pages = 4114–4117|last2 = Hasler|first2 = Paul.|doi = 10.1109/ISCAS.2006.1693534|isbn = 978-0-7803-9389-9|s2cid = 206966013}}</ref> This chip was the first in a line of increasingly complex arrays of floating gate transistors that allowed programmability of charge on the gates of [[MOSFET]]s to model the channel-ion characteristics of neurons in the brain and was one of the first cases of a silicon programmable array of neurons.<br />
<br />
早在2006年，佐治亚理工学院的研究人员就研发出了一种现场可编程神经阵列。<ref name=":12" />在此之后，一系列越来越复杂的浮栅晶体管阵列被成功研发出来，这些晶体管阵列可以通过在'''<font color="#ff8000">金属-氧化物半导体效应晶体管MOSFET</font>'''的栅极上编程来模拟大脑中神经元的离子通道特性，这也是以硅为主要材料的可编程神经元阵列的最早成功案例之一。<br />
<br />
In November 2011, a group of [[MIT]] researchers created a computer chip that mimics the analog, ion-based communication in a synapse between two neurons using 400 transistors and standard [[CMOS]] manufacturing techniques.<ref name=":13">{{cite web|title=MIT creates "brain chip"|url=http://www.extremetech.com/extreme/105067-mit-creates-brain-chip|access-date=4 December 2012}}</ref><ref name="Neuromorphic silicon paper">{{cite journal|title=Neuromorphic silicon neurons and large-scale neural networks: challenges and opportunities|doi=10.3389/fnins.2011.00108|pmid=21991244|pmc=3181466|volume=5|pages=108|journal=Frontiers in Neuroscience|year=2011|last1=Poon|first1=Chi-Sang|last2=Zhou|first2=Kuan|doi-access=free}}</ref><br />
<br />
2011年11月，麻省理工学院的一组研究人员研发出一种计算机芯片，该芯片上使用标准的'''<font color="#ff8000">互补金属氧化物半导体CMOS</font>'''制造技术集成了400个晶体管，能够模拟神经元间突触中基于离子的通讯。<ref name=":13" /><ref name="Neuromorphic silicon paper" /><br />
<br />
In June 2012, [[spintronic]] researchers at [[Purdue University]] presented a paper on the design of a neuromorphic chip using [[Spin valve|lateral spin valve]]s and [[memristor]]s. They argue that the architecture works similarly to neurons and can therefore be used to test methods of reproducing the brain's processing. In addition, these chips are significantly more energy-efficient than conventional ones.<ref name="Spin Devices Prop">{{Cite arXiv|title=Proposal For Neuromorphic Hardware Using Spin Devices|eprint=1206.3227|last1=Sharad|first1=Mrigank|last2=Augustine|first2=Charles|last3=Panagopoulos|first3=Georgios|last4=Roy|first4=Kaushik|class=cond-mat.dis-nn|year=2012}}</ref><br />
<br />
2012年6月，普渡大学的'''<font color="#ff8000">自旋电子学Spintronic</font>'''研究人员发表了一篇关于利用'''<font color="#ff8000">侧向自旋阀Lateral spin valves</font>'''和'''<font color="#ff8000">记忆电阻器Memristors</font>'''设计神经形态芯片的论文。他们认为，这种芯片结构的工作原理与神经元相似，因此可以用于大脑运行机制的复刻方法的测试。此外，这些芯片在能耗方面明显优于传统芯片。<ref name="Spin Devices Prop" /><br />
<br />
Research at [[HP Labs]] on Mott memristors has shown that while they can be non-[[Volatile memory|volatile]], the volatile behavior exhibited at temperatures significantly below the [[phase transition]] temperature can be exploited to fabricate a [[neuristor]],<ref name=":0" /> a biologically-inspired device that mimics behavior found in neurons.<ref name=":0">{{Cite journal | doi = 10.1038/nmat3510| pmid = 23241533| title = A scalable neuristor built with Mott memristors| journal = Nature Materials| volume = 12| issue = 2| pages = 114–7| year = 2012| last1 = Pickett | first1 = M. D. | last2 = Medeiros-Ribeiro | first2 = G. | last3 = Williams | first3 = R. S. | bibcode = 2013NatMa..12..114P| s2cid = 16271627| url = https://semanticscholar.org/paper/b6ba6f496ace2b947f111059663e76bb60e9efeb}}</ref> In September 2013, they presented models and simulations that show how the spiking behavior of these neuristors can be used to form the components required for a [[Turing machine]].<ref name=":14">{{cite journal|doi=10.1088/0957-4484/24/38/384002|title=Phase transitions enable computational universality in neuristor-based cellular automata|author1=Matthew D Pickett|author2=R Stanley Williams|name-list-style=amp|date=September 2013|publisher=IOP Publishing Ltd|journal=Nanotechnology|volume=24|issue=38|pmid=23999059|bibcode=2013Nanot..24L4002P|at=384002}}</ref><br />
<br />
'''<font color="#ff8000">惠普实验室HP labs</font>'''在莫特记忆电阻器上的研究表明，尽管它们可以是非'''<font color="#ff8000">易失性Volatile</font>'''的，但是在'''<font color="#ff8000">相变Phase transition</font>'''温度以下时表现出的易失性行为可以被用来制造'''<font color="#ff8000">类神经元电阻器Neuristor</font>'''(一种生物学启发的模仿神经元行为的硬件)<ref name=":0" />。2013年9月，他们通过模型和仿真展示了这些类神经元电阻器的脉冲行为如何产生'''<font color="#ff8000">图灵机Turing machine</font>'''的所需元素。<ref name=":14" /><br />
<br />
[[Neurogrid]], built by ''Brains in Silicon'' at [[Stanford University]],<ref name=":15">{{cite journal|last1=Boahen|first1=Kwabena|title=Neurogrid: A Mixed-Analog-Digital Multichip System for Large-Scale Neural Simulations|journal=Proceedings of the IEEE|date=24 April 2014|volume=102|issue=5|pages=699–716|doi=10.1109/JPROC.2014.2313565|s2cid=17176371}}</ref> is an example of hardware designed using neuromorphic engineering principles. The circuit board is composed of 16 custom-designed chips, referred to as NeuroCores. Each NeuroCore's analog circuitry is designed to emulate neural elements for 65536 neurons, maximizing energy efficiency. The emulated neurons are connected using digital circuitry designed to maximize spiking throughput.<ref name=":16">{{cite journal|doi=10.1038/503022a|pmid = 24201264|title = Neuroelectronics: Smart connections|journal = Nature|volume = 503|issue = 7474|pages = 22–4|year = 2013|last1 = Waldrop|first1 = M. Mitchell|bibcode = 2013Natur.503...22W|doi-access = free}}</ref><ref name=":17">{{cite journal|doi=10.1109/JPROC.2014.2313565|title = Neurogrid: A Mixed-Analog-Digital Multichip System for Large-Scale Neural Simulations|journal = Proceedings of the IEEE|volume = 102|issue = 5|pages = 699–716|year = 2014|last1 = Benjamin|first1 = Ben Varkey|last2 = Peiran Gao|last3 = McQuinn|first3 = Emmett|last4 = Choudhary|first4 = Swadesh|last5 = Chandrasekaran|first5 = Anand R.|last6 = Bussat|first6 = Jean-Marie|last7 = Alvarez-Icaza|first7 = Rodrigo|last8 = Arthur|first8 = John V.|last9 = Merolla|first9 = Paul A.|last10 = Boahen|first10 = Kwabena|s2cid = 17176371}}</ref><br />
<br />
'''<font color="#ff8000">神经栅格Neurogrid<ref name=":15" /></font>'''是由斯坦福大学Brains in Silicon公司研发的、使用神经形态工程原理设计的硬件。该电路板由16个定制设计的芯片组成(NeuroCores)。在设计中，每个NeuroCore芯片的模拟电路对65536个神经元的神经元素进行模拟，以最大限度地提高能量效率。模拟出的神经元通过设计的数字电路连接，以最大化脉冲吞吐量。<ref name=":16" /><ref name=":17" /><br />
<br />
A research project with implications for neuromorphic engineering is the [[Human Brain Project]] that is attempting to simulate a complete human brain in a supercomputer using biological data. It is made up of a group of researchers in neuroscience, medicine, and computing.<ref name=":18">{{cite web|title=Involved Organizations|url=http://www.humanbrainproject.eu/partners.html|access-date=22 February 2013|url-status=dead|archive-url=https://web.archive.org/web/20130302142627/http://www.humanbrainproject.eu/partners.html|archive-date=2 March 2013}}</ref> [[Henry Markram]], the project's co-director, has stated that the project proposes to establish a foundation to explore and understand the brain and its diseases, and to use that knowledge to build new computing technologies. The three primary goals of the project are to better understand how the pieces of the brain fit and work together, to understand how to objectively diagnose and treat brain diseases, and to use the understanding of the human brain to develop neuromorphic computers. That the simulation of a complete human brain will require a supercomputer a thousand times more powerful than today's encourages the current focus on neuromorphic computers.<ref name=":19">{{cite web|title=Human Brain Project|url=http://www.humanbrainproject.eu|access-date=22 February 2013}}</ref> $1.3 billion has been allocated to the project by The [[European Commission]].<ref name=":20">{{cite web|title=The Human Brain Project and Recruiting More Cyberwarriors|url=http://www.marketplace.org/topics/tech/human-brain-project-and-recruiting-more-cyberwarriors|access-date=22 February 2013|date=January 29, 2013}}</ref><br />
<br />
'''<font color="#ff8000">人类大脑计划Human Brain Project</font>'''对神经形态工程具有较大影响，其主要任务是尝试用生物数据在超级计算机中模拟完整的人脑。人类大脑计划由神经科学、医学和计算机科学背景的研究人员组成。<ref name=":18" />该项目的联合主管亨利•马克拉姆(Henry Markram)表示，人类大脑计划的目的是建立一个探索和了解脑科学和脑疾病知识的基础，并利用这些知识来构建更先进的计算机技术。这个项目的三个主要目标分别是: 更好地理解大脑的各个部分是如何相互配合协同工作的; 理解如何客观地诊断和治疗脑部疾病; 以及利用对人类大脑的理解来开发神经形态计算机。模拟一个完整的人类大脑需要一台比现在强大一千倍的超级计算机，这不断激发着对神经形态计算机领域的研究兴趣。<ref name=":19" />欧盟委员会已经向人类大脑计划拨款13亿美元。<ref name=":20" /><br />
<br />
Other research with implications for neuromorphic engineering involves the [[BRAIN Initiative]]<ref name="economist">[https://www.economist.com/news/science-and-technology/21582495-computers-will-help-people-understand-brains-better-and-understanding-brains Neuromorphic computing: The machine of a new soul], The Economist, 2013-08-03</ref> and the [[TrueNorth]] chip from [[IBM]].<ref name=":21">{{cite journal|last1=Modha|first1=Dharmendra|title=A million spiking-neuron integrated circuit with a scalable communication network and interface|journal=Science|date=Aug 2014|volume=345|issue=6197|pages=668–673|doi=10.1126/science.1254642|pmid=25104385|bibcode=2014Sci...345..668M|s2cid=12706847}}</ref> Neuromorphic devices have also been demonstrated using nanocrystals, nanowires, and conducting polymers.<ref name=":22">{{Cite web|url=http://jessamynfairfield.com/wp-content/uploads/2017/03/PWMar17Fairfield.pdf|title=Smarter Machines|last=Fairfield|first=Jessamyn|date=March 1, 2017}}</ref><br />
<br />
其他与神经形态工程有关的研究还包括'''<font color="#ff8000">脑计划BRAIN initiative</font>'''，<ref name="economist" />和IBM研发的'''<font color="#ff8000">TrueNorth</font>'''芯片。<ref name=":21" />使用纳米晶体、纳米线和导电聚合物也能够用于制造神经形态学硬件。<ref name=":22" /><br />
<br />
[[Intel]] unveiled its neuromorphic research chip, called “[[Intel Loihi|Loihi]]”, in October 2017. The chip uses an asynchronous [[spiking neural network]] (SNN) to implement adaptive self-modifying event-driven fine-grained parallel computations used to implement learning and inference with high efficiency.<ref name=":23">{{cite journal |last1=Davies |first1=Mike |title=Loihi: A Neuromorphic Manycore Processor with On-Chip Learning |journal=IEEE Micro |date=January 16, 2018 |volume=38 |issue=1 |pages=82–99 |display-authors=etal|doi=10.1109/MM.2018.112130359 |s2cid=3608458 }}</ref><ref name="Morris2017">{{cite web |last1=Morris |first1=John |title=Why Intel built a neuromorphic chip |url=https://www.zdnet.com/article/why-intel-built-a-neuromorphic-chip/ |website=ZDNet |access-date=17 August 2018 |language=en}}</ref><br />
<br />
2017年10月，英特尔发布了神经形态芯片'''<font color="#ff8000">Loihi</font>'''。该芯片采用异步'''<font color="#ff8000">脉冲神经网络Spiking neural network</font>'''实现了自适应、自修改、事件驱动的细粒度并行计算，实现了高效的学习和推理。<ref name=":23" /><ref name="Morris2017" /><br />
<br />
[[IMEC]], a Belgium-based nanoelectronics research center, demonstrated the world's first self-learning neuromorphic chip. The brain-inspired chip, based on OxRAM technology, has the capability of self-learning and has been demonstrated to have the ability to compose music.<ref name=":24">{{cite web |title=Imec demonstrates self-learning neuromorphic chip that composes music |url=https://www.imec-int.com/en/articles/imec-demonstrates-self-learning-neuromorphic-chip-that-composes-music |website=IMEC International |access-date=1 October 2019}}</ref> IMEC released the 30-second tune composed by the prototype. The chip was sequentially loaded with songs in the same time signature and style. The songs were old Belgian and French flute minuets, from which the chip learned the rules at play and then applied them.<ref name=":25">{{cite web|last1=Bourzac|first1=Katherine|title=A Neuromorphic Chip That Makes Music|url=https://spectrum.ieee.org/a-neuromorphic-chip-that-makes-music|url-status=live|access-date=1 October 2019|website=IEEE Spectrum|date=May 23, 2017}}</ref><br />
<br />
比利时的'''<font color="#ff8000">微电子研究中心IMEC</font>'''研发了世界上首个自学习神经形态芯片。这种基于'''<font color="#32CD32"> OxRAM(filamentary-oxide-based resistive memory technology)</font>''' 技术的大脑启发芯片具有自学习能力，并且已被证明具有创作音乐的能力。<ref name=":24" />IMEC发布了由芯片原型机谱写的30秒旋律。向芯片加载一系列特征、风格统一的歌曲（古代比利时和法国长笛小步舞曲），芯片从中学习相关规则并将其应用于创作。<ref name=":25" /><br />
<br />
[[Blue Brain Project|The Blue Brain Project]], led by Henry Markram, aims to build biologically detailed digital reconstructions and simulations of the mouse brain. The Blue Brain Project has created in silico models of rodent brains, while attempting to replicate as many details about its biology as possible. The supercomputer-based simulations offer new perspectives on understanding the structure and functions of the brain.<br />
<br />
由亨利·马克拉姆领导的'''<font color="#ff8000">蓝脑计划The Blue Brain Project</font>'''旨在建立小鼠大脑生理学细节的数字重建和模拟。蓝脑计划已经建立了啮齿动物大脑的电子模型，同时进行着尽可能多地复制其生理学细节的尝试。基于超级计算机的模拟为理解大脑的结构和功能提供了新的视角。<br />
<br />
The European Union funded a series of projects at the University of Heidelberg, which led to the development of <br />
[[BrainScaleS]] (brain-inspired multiscale computation in neuromorphic hybrid systems), a hybrid analog [[neuromorphic]] supercomputer located at Heidelberg University, Germany. It was developed as part of the [[Human Brain Project]] neuromorphic computing platform and is the complement to the [[SpiNNaker]] supercomputer (which is based on digital technology). The architecture used in BrainScaleS mimics biological neurons and their connections on a physical level; additionally, since the components are made of silicon, these model neurons operate on average 864 times (24 hours of real time is 100 seconds in the machine simulation) that of their biological counterparts.<ref name=":26">{{Cite web|date=2016-03-21|title=Beyond von Neumann, Neuromorphic Computing Steadily Advances|url=https://www.hpcwire.com/2016/03/21/lacking-breakthrough-neuromorphic-computing-steadily-advance/|access-date=2021-10-08|website=HPCwire|language=en-US}}</ref><br />
<br />
欧盟资助了海德堡大学一系列促进BrainScaleS(神经形态混合系统中受大脑启发的多尺度计算)发展的项目，这是一台位于德国海德堡大学的混合模拟'''<font color="#ff8000">神经形态Neuromorphic</font>'''超级计算机。它是作为人类大脑计划中神经形态计算平台的一部分而开发的，是'''<font color="#ff8000">SpiNNaker</font>'''超级计算机(基于数字技术)的补充。BrainScaleS中使用的体系架构模拟了生物神经元及其在物理层面上的连接；此外，由于这些组件是由硅制成的，这些模型神经元平均运行速度是生物神经元的864倍，这意味着在机器模拟中，24小时的实时时间仅为100秒。<ref name=":26" /><br />
<br />
===Neuromorphic sensors===<br />
===神经形态传感器===<br />
The concept of neuromorphic systems can be extended to sensors (not just to computation). An example of this applied to detecting [[light]] is the [[retinomorphic sensor]] or, when employed in an array, the [[event camera]].<br />
<br />
神经形态系统的概念可以扩展到传感器(而不仅仅是计算单元)。一个用于检测光线的例子是'''<font color="#ff8000">类视网膜传感器Retinomorphic sensor</font>'''，或者'''<font color="#ff8000">事件摄像机Event camera</font>阵列。'''<br />
<br />
<br />
==Ethical considerations ==<br />
==伦理问题==<br />
While the interdisciplinary concept of neuromorphic engineering is relatively new, many of the same ethical considerations apply to neuromorphic systems as apply to [[human-like machines]] and [[artificial intelligence]] in general. However, the fact that neuromorphic systems are designed to mimic a [[human brain]] gives rise to unique ethical questions surrounding their usage.<br />
<br />
虽然神经形态工程这个跨学科概念相对较新，但许多适用于'''<font color="#ff8000">类人机器Human-like machines</font>'''和人工智能的伦理考虑大体上也适用于神经形态系统。另一方面，神经形态系统是为了模仿人类大脑而设计，这一事实引起了一些针对性的独特伦理问题。<br />
<br />
However, the practical debate is that neuromorphic hardware as well as artificial "neural networks" are immensely simplified models of how the brain operates or processes information at a much lower [[complex system|complexity in terms of size and functional technology]] and a much more regular structure in terms of [[brain connectivity|connectivity]]. Comparing [[neuromorphic chip]]s to the brain is a very crude comparison similar to comparing a plane to a bird just because they both have wings and a tail. The fact is that neural cognitive systems are many orders of magnitude more [[Energy efficiency (physics)|energy-]] and compute-efficient than current state-of-the-art AI and neuromorphic engineering is an attempt to narrow this gap by inspiring from the brain's mechanism just like many engineering designs have [[bioengineering|bio-inspired features]].<br />
<br />
然而，这种争论的实际情况是，神经形态硬件和人工“神经网络”是对大脑运作或处理信息过程的极其简化的模型，在大小和功能技术方面的'''<font color="#ff8000">复杂性Complexity</font>'''相比而言比较低，在'''<font color="#ff8000">连接性Connectivity</font>'''方面也具有更加规则的结构。将'''<font color="#ff8000">神经形态芯片Neuromorphic chips</font>'''与大脑进行比较是一种非常粗糙的比较，类似于仅仅因为飞机有翅膀和尾巴就将它与鸟进行比较。事实上，当前最先进的人工智能在'''<font color="#ff8000">能耗效率Energy-efficiency</font>'''和计算效率方面距离人脑神经认知系统仍有较大差距，而神经形态工程只是一种通过从大脑机制中获得灵感来缩小这种差距的尝试，就像许多工程设计中都具有'''<font color="#ff8000">生物启发的特征Bio-inspired features</font>'''一样。<br />
<br />
===Democratic concerns===<br />
===公众担忧===<br />
Significant ethical limitations may be placed on neuromorphic engineering due to public perception.<ref name=":27">{{Cite report|url=https://ai100.stanford.edu/sites/g/files/sbiybj9861/f/ai_100_report_0831fnl.pdf|title=Artificial Intelligence and Life in 2030|author=2015 Study Panel|date=September 2016|work=One Hundred Year Study on Artificial Intelligence (AI100)|publisher=Stanford University}}</ref> Special [[Eurobarometer]] 382: Public Attitudes Towards Robots, a survey conducted by the European Commission, found that 60% of [[European Union]] citizens wanted a ban of robots in the care of children, the elderly, or the disabled. Furthermore, 34% were in favor of a ban on robots in education, 27% in healthcare, and 20% in leisure. The European Commission classifies these areas as notably “human.” The report cites increased public concern with robots that are able to mimic or replicate human functions. Neuromorphic engineering, by definition, is designed to replicate the function of the human brain.<ref name=":1">{{Cite web|url=http://ec.europa.eu/commfrontoffice/publicopinion/archives/ebs/ebs_382_en.pdf|title=Special Eurobarometer 382: Public Attitudes Towards Robots|last=European Commission|date=September 2012|website=European Commission}}</ref><br />
<br />
由于公众认知的相关忧虑，神经形态工程学可能会受到严重的伦理限制。<ref name=":27" />欧盟委员会进行的一项调查发现，60% 的欧盟公民希望禁止机器人参与照顾儿童、老人或残疾人的工作。此外，34% 的人支持禁止机器人用于教育，27% 的人支持禁止机器人用于医疗保健，20% 的人支持禁止机器人用于休闲。欧盟委员会将以上领域划入“人类”范畴。报告指出，公众越来越关注能够模仿或复制人类功能的机器人。神经形态工程，顾名思义，是为了复制人脑的功能而设计的。<ref name=":1" /><br />
<br />
The democratic concerns surrounding neuromorphic engineering are likely to become even more profound in the future. The European Commission found that EU citizens between the ages of 15 and 24 are more likely to think of robots as human-like (as opposed to instrument-like) than EU citizens over the age of 55. When presented an image of a robot that had been defined as human-like, 75% of EU citizens aged 15–24 said it corresponded with the idea they had of robots while only 57% of EU citizens over the age of 55 responded the same way. The human-like nature of neuromorphic systems, therefore, could place them in the categories of robots many EU citizens would like to see banned in the future.<ref name=":1" /><br />
<br />
围绕神经形态工程的公众担忧可能在未来变得更加深刻。欧盟委员会发现，相比于55岁以上的欧盟公民，15至24岁的欧盟公民更有可能认为机器人像人(而不是像仪器)。当看到一张“类人”机器人的图片时，年龄在15岁至24岁之间的欧盟公民中有75% 的人表示这符合他们对机器人的想法，而55岁以上的欧盟公民中只有57% 的人有同样的反应。因此，神经形态系统可能因为其类似人类的特性而被归入许多欧盟公民希望在未来禁止使用的机器人类别。<ref name=":1" /><br />
<br />
=== Personhood===<br />
=== 人格问题===<br />
As neuromorphic systems have become increasingly advanced, some scholars{{who|date=August 2021}} have advocated for granting [[personhood]] rights to these systems. If the brain is what grants humans their personhood, to what extent does a neuromorphic system have to mimic the human brain to be granted personhood rights? Critics of technology development in the [[Human Brain Project]], which aims to advance brain-inspired computing, have argued that advancement in neuromorphic computing could lead to machine consciousness or personhood.<ref name=":28">{{Cite journal|last=Aicardi|first=Christine|date=September 2018|title=Accompanying technology development in the Human Brain Project: From foresight to ethics management|journal=Futures|volume=102|pages=114–124|doi=10.1016/j.futures.2018.01.005|doi-access=free}}</ref> If these systems are to be treated as people, critics argue, then many tasks humans perform using neuromorphic systems, including the act of termination of neuromorphic systems, may be morally impermissible as these acts would violate the autonomy of the neuromorphic systems.<ref name=":29">{{Cite journal|last=Lim|first=Daniel|date=2014-06-01|title=Brain simulation and personhood: a concern with the Human Brain Project|journal=Ethics and Information Technology|language=en|volume=16|issue=2|pages=77–89|doi=10.1007/s10676-013-9330-5|s2cid=17415814|issn=1572-8439}}</ref><br />
<br />
随着神经形态系统的日益发展，一些学者主张赋予这些系统'''<font color="ff8000">人格Personhood</font>'''权。如果是大脑赋予了人类人格，那么在多大程度上模仿人类大脑的神经形态系统才能被赋予人格权利？“人类大脑计划”旨在推进以大脑为灵感的计算机技术发展，该计划的批评者认为，神经形态计算机技术的进步可能导致机器意识或人格的形成。<ref name=":28" />这些批评者认为，如果这些系统被当作人来对待，那么人类使用神经形态系统执行任务（包括终止神经形态系统）的行为，在道德上就可能是不被允许的，因为这些行为将违反神经形态系统的自主性。<ref name=":29" /><br />
<br />
<br />
==Dual use (military applications)==<br />
==军民两用技术（军事应用）==<br />
The [[Joint Artificial Intelligence Center]], a branch of the U.S. military, is a center dedicated to the procurement and implementation of AI software and neuromorphic hardware for combat use. Specific applications include smart headsets/goggles and robots. JAIC intends to rely heavily on neuromorphic technology to connect "every fighter every shooter" within a network of neuromorphic-enabled units.<br />
<br />
'''<font color="#ff8000">联合人工智能中心The Joint Artificial Intelligence Center</font>'''(JAIC),是美国军队的一个分支，专门从事采购和实施用于战斗的人工智能软件和神经形态硬件。具体应用包括智能耳机/护目镜和机器人。JAIC打算高度依赖神经形态技术，'''<font color="#32CD32">使用神经形态技术来连接神经形态单位网络中的“每个战士每个射手”</font>'''。<br />
<br />
<br />
==Legal considerations==<br />
==法律问题==<br />
Skeptics have argued that there is no way to apply the electronic personhood, the concept of personhood that would apply to neuromorphic technology, legally. In a letter signed by 285 experts in law, robotics, medicine, and ethics opposing a European Commission proposal to recognize “smart robots” as legal persons, the authors write, “A legal status for a robot can’t derive from the [[Natural person|Natural Person]] model, since the robot would then hold [[human rights]], such as the right to dignity, the right to its integrity, the right to remuneration or the right to citizenship, thus directly confronting the Human rights. This would be in contradiction with the [[Charter of Fundamental Rights of the European Union]] and the [[Convention for the Protection of Human Rights and Fundamental Freedoms]].”<ref name=":30">{{Cite web|url=http://www.robotics-openletter.eu/|title=Robotics Openletter {{!}} Open letter to the European Commission|language=fr-FR|access-date=2019-05-10}}</ref><br />
<br />
质疑者者认为，在法律上没有办法应用能够适用于神经形态技术的电子人格。在一封由285名法律、机器人技术、医学和伦理学专家的联名信中，作者们反对欧盟委员会提出的承认“智能机器人”为法人的提议。他们写道，“机器人的法律地位无法从'''<font color="#ff8000">自然人Natural Person</font>'''模型中推导出来，因为机器人将被赋予'''<font color="#ff8000">人权Human rights</font>'''，如尊严权、完整权、报酬权或公民权，从而直接面临人权问题。这将有悖于《欧联基本权利宪章》和《欧洲保障人权和根本自由公约》”。<ref name=":30" /><br />
<br />
===Ownership and property rights ===<br />
===所有权及财产权问题===<br />
There is significant legal debate around property rights and artificial intelligence. In ''Acohs Pty Ltd v. Ucorp Pty Ltd'', Justice Christopher Jessup of the [[Federal Court of Australia]] found that the [[source code]] for [[Material safety data sheets|Material Safety Data Sheets]] could not be [[Copyright law of Australia|copyrighted]] as it was generated by a [[software interface]] rather than a human author.<ref name=":31">{{Cite web|url=http://www.lavan.com.au/advice/intellectual_property_technology/copyright_in_source_code_and_digital_products|title=Copyright in source code and digital products|last=Lavan|website=Lavan|language=en|access-date=2019-05-10}}</ref> The same question may apply to neuromorphic systems: if a neuromorphic system successfully mimics a human brain and produces a piece of original work, who, if anyone, should be able to claim ownership of the work?<ref name=":32">{{cite journal |last1=Eshraghian|first1=Jason K. |title=Human Ownership of Artificial Creativity |journal=Nature Machine Intelligence |date=9 March 2020 |volume=2 |pages=157–160 |doi=10.1038/s42256-020-0161-x}}</ref><br />
<br />
法律界围绕财产权和人工智能有着重大争论。在Acohs Pty有限公司诉Ucorp Pty有限公司一案中，澳大利亚联邦法院的克里斯托弗·杰瑟普法官发现，版权保护不适用于材料安全数据表的'''<font color="#ff8000">源代码Source code</font>'''，因为它是由'''<font color="#ff8000">软件界面Software interface</font>'''生成而非人类工作者生成的。<ref name=":31" />同样的问题可能也适用于神经形态系统: 如果一个神经形态系统成功地模仿了人类的大脑，并产生了一部原创作品，那么该如何确认这部作品的所有权归属？<ref name=":32" /><br />
<br />
<br />
==Neuromemristive systems==<br />
==神经记忆电阻系统==<br />
Neuromemristive systems are a subclass of neuromorphic computing systems that focus on the use of [[memristors]] to implement [[neuroplasticity]]. While neuromorphic engineering focuses on mimicking biological behavior, neuromemristive systems focus on abstraction.<ref>{{Cite web|url=https://digitalops.sandia.gov/Mediasite/Play/a10cf6ceb55d47608bb8326dd00e46611d|title=002.08 N.I.C.E. Workshop 2014: Towards Intelligent Computing with Neuromemristive Circuits and Systems - Feb. 2014|website=digitalops.sandia.gov|access-date=2019-08-26}}</ref> For example, a neuromemristive system may replace the details of a [[Cerebral cortex|cortical]] microcircuit's behavior with an abstract neural network model.<ref>C. Merkel and D. Kudithipudi, "Neuromemristive extreme learning machines for pattern classification," ISVLSI, 2014.</ref><br />
<br />
Neuromemristive systems are a subclass of neuromorphic computing systems that focus on the use of memristors to implement neuroplasticity. While neuromorphic engineering focuses on mimicking biological behavior, neuromemristive systems focus on abstraction. For example, a neuromemristive system may replace the details of a cortical microcircuit's behavior with an abstract neural network model.C. Merkel and D. Kudithipudi, "Neuromemristive extreme learning machines for pattern classification," ISVLSI, 2014.<br />
<br />
神经记忆电阻系统是神经形态计算系统的一个亚类，主要研究利用记忆电阻器实现神经可塑性。神经形态工程的重点是模拟生物行为，而神经记忆电阻系统的重点是提取。例如，一个神经记忆系统可以用一个抽象的神经网络模型取代皮层微电路的行为细节。默克尔和 d. Kudithipudi，“用于模式分类的神经记忆极端学习机器”，ISVLSI，2014。<br />
<br />
There exist several neuron inspired threshold logic functions<ref name="Maan 1–13" /> implemented with memristors that have applications in high level [[pattern recognition]] applications. Some of the applications reported recently include [[speech recognition]],<ref>{{Cite journal|title = Memristor pattern recogniser: isolated speech word recognition|journal = Electronics Letters|pages = 1370–1372|volume = 51|issue = 17|doi = 10.1049/el.2015.1428|first1 = A.K.|last1 = Maan|first2 = A.P.|last2 = James|first3 = S.|last3 = Dimitrijev|year = 2015|bibcode = 2015ElL....51.1370M|hdl = 10072/140989|s2cid = 61454815|url = https://semanticscholar.org/paper/48d3ab11ec6e213b62f11eedcfb7b7febb058674|hdl-access = free}}</ref> [[face recognition]]<ref>{{Cite journal|title = Memristive Threshold Logic Face Recognition|journal = Procedia Computer Science|date = 2014-01-01|pages = 98–103|volume = 41|series = 5th Annual International Conference on Biologically Inspired Cognitive Architectures, 2014 BICA|doi = 10.1016/j.procs.2014.11.090|first1 = Akshay Kumar|last1 = Maan|first2 = Dinesh S.|last2 = Kumar|first3 = Alex Pappachen|last3 = James|doi-access = free}}</ref> and [[object recognition]].<ref>{{Cite journal|title = Memristive Threshold Logic Circuit Design of Fast Moving Object Detection|journal = IEEE Transactions on Very Large Scale Integration (VLSI) Systems|date = 2015-10-01|issn = 1063-8210|pages = 2337–2341|volume = 23|issue = 10|doi = 10.1109/TVLSI.2014.2359801|first1 = A.K.|last1 = Maan|first2 = D.S.|last2 = Kumar|first3 = S.|last3 = Sugathan|first4 = A.P.|last4 = James|arxiv = 1410.1267|s2cid = 9647290}}</ref> They also find applications in replacing conventional digital logic gates.<ref>{{Cite journal|title = Resistive Threshold Logic|journal = IEEE Transactions on Very Large Scale Integration (VLSI) Systems|date = 2014-01-01|issn = 1063-8210|pages = 190–195|volume = 22|issue = 1|doi = 10.1109/TVLSI.2012.2232946|first1 = A.P.|last1 = James|first2 = L.R.V.J.|last2 = Francis|first3 = D.S.|last3 = Kumar|arxiv = 1308.0090|s2cid = 7357110}}</ref><ref>{{Cite journal|title = Threshold Logic Computing: Memristive-CMOS Circuits for Fast Fourier Transform and Vedic Multiplication|journal = IEEE Transactions on Very Large Scale Integration (VLSI) Systems|date = 2015-11-01|issn = 1063-8210|pages = 2690–2694|volume = 23|issue = 11|doi = 10.1109/TVLSI.2014.2371857|first1 = A.P.|last1 = James|first2 = D.S.|last2 = Kumar|first3 = A.|last3 = Ajayan|arxiv = 1411.5255|s2cid = 6076956}}</ref><br />
<br />
There exist several neuron inspired threshold logic functions implemented with memristors that have applications in high level pattern recognition applications. Some of the applications reported recently include speech recognition, face recognition and object recognition. They also find applications in replacing conventional digital logic gates.<br />
<br />
利用记忆电阻器实现的神经元启发阈值逻辑函数在高级模式识别中有着广泛的应用。最近报道的一些应用包括语音识别、人脸识别和物体识别。它们还可以用来取代传统的数字逻辑门。<br />
<br />
For ideal passive memristive circuits there is an exact equation (Caravelli-Traversa-[[Di Ventra]] equation) for the internal memory of the circuit:<ref>{{cite journal |last=Caravelli |display-authors=etal|arxiv=1608.08651 |title=The complex dynamics of memristive circuits: analytical results and universal slow relaxation |year=2017 |doi=10.1103/PhysRevE.95.022140 |pmid= 28297937 |volume=95 |issue= 2 |pages= 022140 |journal=Physical Review E|bibcode=2017PhRvE..95b2140C |s2cid=6758362}}</ref><br />
<br />
For ideal passive memristive circuits there is an exact equation (Caravelli-Traversa-Di Ventra equation) for the internal memory of the circuit:<br />
<br />
对于理想的无源记忆电路，电路的内部记忆有一个精确的方程(Caravelli-Traversa-Di Ventra 方程) :<br />
<br />
:<math> \frac{d}{dt} \vec{W} = \alpha \vec{W}-\frac{1}{\beta} (I+\xi \Omega W)^{-1} \Omega \vec S </math><br />
<br />
:\frac{d}{dt} \vec{W} = \alpha \vec{W}-\frac{1}{\beta} (I+\xi \Omega W)^{-1} \Omega \vec S<br />
<br />
:\frac{d}{dt} \vec{W} = \alpha \vec{W}-\frac{1}{\beta} (I+\xi \Omega W)^{-1} \Omega \vec S<br />
<br />
as a function of the properties of the physical memristive network and the external sources. In the equation above, <math>\alpha</math> is the "forgetting" time scale constant, <math> \xi=r-1</math> and <math>r=\frac{R_\text{off}}{R_\text{on}}</math> is the ratio of ''off'' and ''on'' values of the limit resistances of the memristors, <math> \vec S </math> is the vector of the sources of the circuit and <math>\Omega</math> is a projector on the fundamental loops of the circuit. The constant <math>\beta</math> has the dimension of a voltage and is associated to the properties of the [[memristor]]; its physical origin is the charge mobility in the conductor. The diagonal matrix and vector <math>W=\operatorname{diag}(\vec W)</math> and <math>\vec W</math> respectively, are instead the internal value of the memristors, with values between 0 and 1. This equation thus requires adding extra constraints on the memory values in order to be reliable.<br />
<br />
<nowiki>as a function of the properties of the physical memristive network and the external sources. In the equation above, \alpha is the "forgetting" time scale constant, \xi=r-1 and r=\frac{R_\text{off}}{R_\text{on}} is the ratio of off and on values of the limit resistances of the memristors, \vec S is the vector of the sources of the circuit and \Omega is a projector on the fundamental loops of the circuit. The constant \beta has the dimension of a voltage and is associated to the properties of the memristor; its physical origin is the charge mobility in the conductor. The diagonal matrix and vector W=\operatorname{diag}(\vec W) and \vec W respectively, are instead the internal value of the memristors, with values between 0 and 1. This equation thus requires adding extra constraints on the memory values in order to be reliable.</nowiki><br />
<br />
<nowiki>作为物理记忆网络和外部源的性质的函数。在上述方程中，α 是“遗忘”时间尺度常数，xi = r-1，r = frac { r _ text { off }{ on }{ r _ text { on }}是记忆电阻器极限电阻的开关和开关值之比，vec s 是电路源的矢量，Omega 是电路基本环路的投影仪。常数 β 具有电压的尺寸，与记忆电阻器的特性有关; 它的物理起源是导体中的电荷迁移率。对角矩阵和向量 w = 操作者名{ diag }(vec w)和 vec w 分别是记忆电阻器的内值，值在0到1之间。因此，这个等式需要在内存值上添加额外的约束，以保证可靠性。</nowiki><br />
<br />
== See also==<br />
==相关词条==<br />
{{Columns-list|colwidth=18em|<br />
* [[AI accelerator (computer hardware)|AI accelerator]]<br />
* [[Artificial brain]]<br />
* [[Biomorphic]]<br />
* [[Cognitive computer]]<br />
* [[Computation and Neural Systems]]<br />
* [[Differentiable programming]]<br />
* [[Event camera]]<br />
* [[Neurorobotics]]<br />
* [[Optical flow sensor]]<br />
* [[Physical neural network]]<br />
* [[SpiNNaker]]<br />
* [[SyNAPSE]]<br />
* [[Retinomorphic sensor]]<br />
* [[Vision chip]]<br />
* [[Vision processing unit]]<br />
* [[Zeroth (software)]]<br />
* [[Hardware for artificial intelligence]]<br />
}}<br />
<br />
{{Portal bar|Electronics}}<br />
<br />
== References ==<br />
{{Reflist|40em}}<br />
<br />
==External links==<br />
<!--======================== {{No more links}} ============================<br />
| PLEASE BE CAUTIOUS IN ADDING MORE LINKS TO THIS ARTICLE. Wikipedia |<br />
| is not a collection of links nor should it be used for advertising. |<br />
| |<br />
| Excessive or inappropriate links WILL BE DELETED. |<br />
| See [[Wikipedia:External links]] & [[Wikipedia:Spam]] for details. |<br />
| |<br />
| If there are already plentiful links, please propose additions or |<br />
| replacements on this article's discussion page, or submit your link |<br />
| to the relevant category at the Open Directory Project (dmoz.org) |<br />
| and link back to that category using the {{dmoz}} template. |<br />
======================={{No more links}}=============================--><br />
{{Commons category|Neuromorphic Engineering}}<br />
*[https://web.archive.org/web/20150727034331/http://ine-web.org/workshops/workshops-overview Telluride Neuromorphic Engineering Workshop]<br />
*[https://archive.today/20130115190057/http://capocaccia.ethz.ch/ CapoCaccia Cognitive Neuromorphic Engineering Workshop]<br />
*[https://web.archive.org/web/20190716132350/http://www.ine-web.org/ Institute of Neuromorphic Engineering]<br />
*[http://www.ine-news.org/ INE news site].<br />
*[http://www.frontiersin.org/neuromorphic_engineering Frontiers in Neuromorphic Engineering Journal]<br />
*[http://www.cns.caltech.edu/ Computation and Neural Systems] department at the [[California Institute of Technology]].<br />
*[http://www.humanbrainproject.eu/ Human Brain Project official site]<br />
*[https://www.the-scientist.com/features/building-a-silicon-brain-65738 Building a Silicon Brain:] Computer chips based on biological neurons may help simulate larger and more-complex brain models. May 1, 2019. SANDEEP RAVINDRAN<br />
<br />
<br />
*Telluride Neuromorphic Engineering Workshop<br />
*CapoCaccia Cognitive Neuromorphic Engineering Workshop<br />
*Institute of Neuromorphic Engineering<br />
*INE news site.<br />
*Frontiers in Neuromorphic Engineering Journal<br />
*Computation and Neural Systems department at the California Institute of Technology.<br />
*Human Brain Project official site<br />
*Building a Silicon Brain: Computer chips based on biological neurons may help simulate larger and more-complex brain models. May 1, 2019. SANDEEP RAVINDRAN<br />
<br />
=<nowiki>外部链接</nowiki>= <br />
* 碲化物神经形态工程工作室<br />
*CapoCaccia 认知神经形态工程工作室<br />
*神经形态工程研究所<br />
*INE 新闻站点。<br />
*《神经形态工程学前沿》<br />
*加州理工学院计算与神经系统系。<br />
*人脑项目官方网站<br />
*打造硅脑: 基于生物神经元的计算机芯片可能有助于模拟更大、更复杂的大脑模型。2019年5月1日。SANDEEP RAVINDRAN<br />
<br />
{{Differentiable computing}}<br />
{{Authority control}}<br />
<br />
[[Category:Electrical engineering]]<br />
[[Category:Neuroscience]]<br />
[[分类:AI accelerators|*]]<br />
[[Category:Artificial intelligence]]<br />
[[Category:Robotics]]<br />
<br />
Category:Electrical engineering<br />
Category:Neuroscience<br />
*<br />
Category:Artificial intelligence<br />
Category:Robotics<br />
<br />
类别: 电气工程类别: 神经科学 * 类别: 人工智能类别: 机器人<br />
<br />
<noinclude><br />
<br />
<small>This page was moved from [[wikipedia:en:Neuromorphic engineering]]. Its edit history can be viewed at [[神经形态计算/edithistory]]</small></noinclude><br />
<br />
[[Category:待整理页面]]<br />
<br />
<references /></div>Shenky20https://wiki.swarma.org/index.php?title=%E7%A5%9E%E7%BB%8F%E5%BD%A2%E6%80%81%E8%AE%A1%E7%AE%97&diff=30748神经形态计算2022-04-28T11:02:52Z<p>Shenky20：/* 法律问题 */</p>
<hr />
<div><br />
此词条由神经动力学模型读书会词条梳理志愿者Shenky20翻译审校，未经专家审校，带来阅读不便，请见谅。<br />
<br />
{{Use American English|date = January 2019}}<br />
{{Use mdy dates|date = January 2019}}<br />
<br />
'''Neuromorphic engineering''', also known as '''neuromorphic computing''',<ref name=":3">{{Cite journal | doi = 10.1145/2601069| title = Neuromorphic computing gets ready for the (really) big time| journal = [[Communications of the ACM]]| volume = 57| issue = 6| year = 2014| last1 = Monroe | first1 = D. | pages = 13–15| s2cid = 20051102}}</ref><ref name=":4">{{Cite journal | doi = 10.1088/0957-4484/21/17/175202| pmid = 20368686| title = Nanotube devices based crossbar architecture: Toward neuromorphic computing| journal = Nanotechnology| volume = 21| issue = 17| pages = 175202| year = 2010| last1 = Zhao | first1 = W. S. | last2 = Agnus | first2 = G. | last3 = Derycke | first3 = V. | last4 = Filoramo | first4 = A. | last5 = Bourgoin | first5 = J. -P. | last6 = Gamrat | first6 = C. | bibcode = 2010Nanot..21q5202Z| url = https://zenodo.org/record/3428659}}</ref><ref name="humanbrainproject">{{YouTube|id=6RoiZ90mGfw|title=The Human Brain Project SP 9: Neuromorphic Computing Platform}}</ref> is the use of [[very-large-scale integration]] (VLSI) systems containing electronic [[analog circuit]]s to mimic neuro-biological architectures present in the nervous system. A neuromorphic computer/chip is any device that uses physical artificial neurons (made from silicon) to do computations.<ref name=":5">{{cite journal|last1=Mead|first1=Carver|title=Neuromorphic electronic systems|journal=Proceedings of the IEEE|date=1990|volume=78|issue=10|pages=1629–1636|doi=10.1109/5.58356|url=https://authors.library.caltech.edu/53090/1/00058356.pdf}}</ref><ref name=":2" /> In recent times, the term ''neuromorphic'' has been used to describe analog, digital, mixed-mode analog/digital VLSI, and software systems that implement models of [[neural system]]s (for [[perception]], [[motor control]], or [[multisensory integration]]). The implementation of neuromorphic computing on the hardware level can be realized by oxide-based [[memristor]]s,<ref name="Maan 1–13">{{Cite journal|last1=Maan|first1=A. K.|last2=Jayadevi|first2=D. A.|last3=James|first3=A. P.|date=2016-01-01|title=A Survey of Memristive Threshold Logic Circuits|journal=IEEE Transactions on Neural Networks and Learning Systems|volume=PP|issue=99|pages=1734–1746|doi=10.1109/TNNLS.2016.2547842|pmid=27164608|issn=2162-237X|arxiv=1604.07121|bibcode=2016arXiv160407121M|s2cid=1798273}}</ref> spintronic memories, threshold switches, and [[transistor]]s.<ref name=":6">{{Cite journal|title = Mott Memory and Neuromorphic Devices|journal = Proceedings of the IEEE|date = 2015-08-01|issn = 0018-9219|pages = 1289–1310|volume = 103|issue = 8|doi = 10.1109/JPROC.2015.2431914|first1 = You|last1 = Zhou|first2 = S.|last2 = Ramanathan|s2cid = 11347598|url = https://zenodo.org/record/895565}}</ref><ref name=":2">{{Cite document|title=Neuromorphic Circuits With Neural Modulation Enhancing the Information Content of Neural Signaling {{!}} International Conference on Neuromorphic Systems 2020|language=EN|doi=10.1145/3407197.3407204|s2cid=220794387}}</ref> Training software-based neuromorphic systems of spiking neural networks can be achieved using error backpropagation, e.g., using [[Python (programming language)|Python]] based frameworks such as snnTorch,<ref name=":7">{{cite journal |last1=Eshraghian|first1=Jason K.|last2=Ward|first2=Max|last3=Neftci |first3=Emre|last4=Wang|first4=Xinxin|last5=Lenz|first5=Gregor|last6=Dwivedi|first6=Girish|last7=Bennamoun|first7=Mohammed|last8=Jeong|first8=Doo Seok|last9=Lu|first9=Wei D.|title=Training Spiking Neural Networks Using Lessons from Deep Learning |date=1 October 2021 |arxiv=2109.12894 }}</ref> or using canonical learning rules from the biological learning literature, e.g., using BindsNet.<ref name=":8">{{Cite web | url=https://github.com/Hananel-Hazan/bindsnet | title=Hananel-Hazan/bindsnet: Simulation of spiking neural networks (SNNs) using PyTorch.| date=31 March 2020}}</ref><br />
<br />
<br />
'''<font color="#ff8000">神经形态工程Neuromorphic engineering</font>'''(也称为'''<font color="#ff8000">神经形态计算Neuromorphic computing</font>''')<ref name=":3" /><ref name=":4" /><ref name="humanbrainproject" />是指使用包含电子'''<font color="#ff8000">模拟电路Analog circuit</font>'''的'''<font color="#ff8000">超大规模集成电路Very-large-scale integration</font>'''系统来模拟神经系统中生理结构的研究方法。神经形态计算机或神经形态芯片包括任何使用由硅制成的人造神经元进行计算的设备。<ref name=":5" /><ref name=":2" />近年来，神经形态学（neuromorphic）这个术语被用来描述能够实现'''<font color="#ff8000">神经系统Neural system</font>'''模型功能(如'''<font color="#ff8000">感知Perception</font>'''、'''<font color="#ff8000">运动控制Motor control</font>'''，'''<font color="#ff8000">多感官整合Multisensory integration</font>'''等)的模拟、数字、模拟/数字混合模式超大规模集成电路和软件系统。神经形态计算的硬件实现可以通过基于氧化物的'''<font color="#ff8000">记忆电阻器Memristor</font>'''、<ref name="Maan 1–13" />自旋电子存储器、阈值开关和'''<font color="#ff8000">晶体管Transistor</font>'''来实现。<ref name=":6" /><ref name=":2" />对基于软件的脉冲神经网络系统的训练可以通过误差反向传播机制来实现，例如，使用snnTorch等基于Python的框架，<ref name=":7" />或使用BindsNet等典型的受生物启发的学习模式。<ref name=":8" /><br />
<br />
A key aspect of neuromorphic engineering is understanding how the morphology of individual neurons, circuits, applications, and overall architectures creates desirable computations, affects how information is represented, influences robustness to damage, incorporates learning and development, adapts to local change (plasticity), and facilitates evolutionary change.<br />
<br />
<br />
神经形态工程领域的一个关键问题，就是理解'''<font color="#32CD32">单个神经元形态、神经回路、应用和整体结构</font>'''如何产生理想的计算，如何影响信息的表示和对破坏的鲁棒性，如何整合学习和发展，如何适应局部变化(可塑性) 并促进逐渐发展的变化。<br />
<br />
Neuromorphic engineering is an interdisciplinary subject that takes inspiration from [[biology]], [[physics]], [[mathematics]], [[computer science]], and [[electronic engineering]]<ref name=":2" /> to design artificial neural systems, such as [[Machine vision|vision systems]], head-eye systems, auditory processors, and autonomous robots, whose physical architecture and design principles are based on those of biological nervous systems.<ref name=":9">{{Cite journal | doi = 10.1155/2012/705483| title = Qualitative Functional Decomposition Analysis of Evolved Neuromorphic Flight Controllers| journal = Applied Computational Intelligence and Soft Computing| volume = 2012| pages = 1–21| year = 2012| last1 = Boddhu | first1 = S. K. | last2 = Gallagher | first2 = J. C. | doi-access = free}}</ref> It was developed by [[Carver Mead]]<ref name=":10">{{cite web|title=carver mead website|last1=Mead|first1=Carver|url=http://www.carvermead.caltech.edu/index.html|website=carvermead}}</ref> in the late 1980s.<br />
<br />
<br />
神经形态工程是以'''<font color="#ff8000">生物学Biology</font>'''、'''<font color="#ff8000">物理学Physics</font>'''、'''<font color="#ff8000">数学Mathematics</font>'''、'''<font color="#ff8000">计算机科学Computer science</font>'''和'''<font color="#ff8000">电子工程Electronic engineering<ref name=":2" /></font>'''等学科为基础，设计人工神经系统（如'''<font color="#ff8000">视觉系统Vision systems</font>'''、头眼系统、听觉处理器以及物理结构和设计原则都受启发于生物神经系统的自主机器人）的一门交叉学科。<ref name=":9" />20世纪80年代后期，卡弗·米德极大地推动了神经形态工程领域的发展。<ref name=":10" /><br />
<br />
==Neurological inspiration==<br />
==神经生物学启发==<br />
Neuromorphic engineering is set apart by the inspiration it takes from what we know about the structure and operations of the [[brain]]. Neuromorphic engineering translates what we know about the brain's function into computer systems. Work has mostly focused on replicating the analog nature of [[biological computation]] and the role of [[neuron]]s in [[cognition]].<br />
<br />
<br />
神经形态工程形成的灵感来源于现已知的'''<font color="#ff8000">大脑Brain</font>'''结构及运作机制知识，它将我们对大脑功能的了解用于研究和优化计算机系统。该领域工作主要集中于对'''<font color="#ff8000">生物计算Biological computation</font>'''模拟特性和'''<font color="#ff8000">神经元Neuron</font>'''在'''<font color="#ff8000">认知Cognition</font>'''中所发挥作用的复刻。<br />
<br />
The biological processes of neurons and their [[synapse]]s are dauntingly complex, and thus very difficult to artificially simulate. A key feature of biological brains is that all of the processing in neurons use analog chemical signals. This makes it hard to replicate brains in computers because the current generation of computers is completely digital. However, the characteristics of these parts can be abstracted into mathematical functions that closely capture the essence of the neuron's operations.<br />
<br />
神经元及其'''<font color="#ff8000">突触Synapse</font>'''的生理学过程极其复杂，难以进行人工模拟。神经元在所有处理过程中都使用化学模拟信号，这是大脑的一个关键生理学特征。这个特征大大增加了在计算机中复制大脑的难度，因为目前的计算机是完全数字化的。然而，部分特征可以抽象为数学函数，这些函数能够紧密捕捉神经元运行的本质。<br />
<br />
The goal of neuromorphic computing is not to perfectly mimic the brain and all of its functions, but instead to extract what is known of its structure and operations to be used in a practical computing system. No neuromorphic system will claim nor attempt to reproduce every element of neurons and synapses, but all adhere to the idea that computation is highly [[distributed processing|distributed]] throughout a series of small computing elements analogous to a neuron. While this sentiment is standard, researchers chase this goal with different methods.<ref name=":11">{{Cite journal | doi = 10.1088/1741-2560/13/5/051001| title = Large-scale neuromorphic computing systems| journal = Journal of Neural Engineering| volume = 13| pages = 1–15| year = 2016| last1 = Furber | first1 = Steve| issue = 5| pmid = 27529195| bibcode = 2016JNEng..13e1001F| doi-access = free}}</ref><br />
<br />
神经形态计算的目标不是完美地模拟大脑及其所有功能，而是利用已知的大脑结构和运转机制来研发或优化实际的计算系统。任何神经形态学系统都不会声称或试图复制神经元和突触中的每一个元素，但所有人都一致认可将计算高度'''<font color="#ff8000">分散Distribute</font>'''于一系列类似于神经元的小型计算元素的理念。研究人员用不同的方法来追求这一普遍目标。<ref name=":11" /><br />
<br />
<br />
==Examples==<br />
==代表性成果==<br />
As early as 2006, researchers at [[Georgia Tech]] published a field programmable neural array.<ref name=":12">{{Cite book|title = A field programmable neural array|last1 = Farquhar|first1 = Ethan|date = May 2006|journal = IEEE International Symposium on Circuits and Systems|pages = 4114–4117|last2 = Hasler|first2 = Paul.|doi = 10.1109/ISCAS.2006.1693534|isbn = 978-0-7803-9389-9|s2cid = 206966013}}</ref> This chip was the first in a line of increasingly complex arrays of floating gate transistors that allowed programmability of charge on the gates of [[MOSFET]]s to model the channel-ion characteristics of neurons in the brain and was one of the first cases of a silicon programmable array of neurons.<br />
<br />
早在2006年，佐治亚理工学院的研究人员就研发出了一种现场可编程神经阵列。<ref name=":12" />在此之后，一系列越来越复杂的浮栅晶体管阵列被成功研发出来，这些晶体管阵列可以通过在'''<font color="#ff8000">金属-氧化物半导体效应晶体管MOSFET</font>'''的栅极上编程来模拟大脑中神经元的离子通道特性，这也是以硅为主要材料的可编程神经元阵列的最早成功案例之一。<br />
<br />
In November 2011, a group of [[MIT]] researchers created a computer chip that mimics the analog, ion-based communication in a synapse between two neurons using 400 transistors and standard [[CMOS]] manufacturing techniques.<ref name=":13">{{cite web|title=MIT creates "brain chip"|url=http://www.extremetech.com/extreme/105067-mit-creates-brain-chip|access-date=4 December 2012}}</ref><ref name="Neuromorphic silicon paper">{{cite journal|title=Neuromorphic silicon neurons and large-scale neural networks: challenges and opportunities|doi=10.3389/fnins.2011.00108|pmid=21991244|pmc=3181466|volume=5|pages=108|journal=Frontiers in Neuroscience|year=2011|last1=Poon|first1=Chi-Sang|last2=Zhou|first2=Kuan|doi-access=free}}</ref><br />
<br />
2011年11月，麻省理工学院的一组研究人员研发出一种计算机芯片，该芯片上使用标准的'''<font color="#ff8000">互补金属氧化物半导体CMOS</font>'''制造技术集成了400个晶体管，能够模拟神经元间突触中基于离子的通讯。<ref name=":13" /><ref name="Neuromorphic silicon paper" /><br />
<br />
In June 2012, [[spintronic]] researchers at [[Purdue University]] presented a paper on the design of a neuromorphic chip using [[Spin valve|lateral spin valve]]s and [[memristor]]s. They argue that the architecture works similarly to neurons and can therefore be used to test methods of reproducing the brain's processing. In addition, these chips are significantly more energy-efficient than conventional ones.<ref name="Spin Devices Prop">{{Cite arXiv|title=Proposal For Neuromorphic Hardware Using Spin Devices|eprint=1206.3227|last1=Sharad|first1=Mrigank|last2=Augustine|first2=Charles|last3=Panagopoulos|first3=Georgios|last4=Roy|first4=Kaushik|class=cond-mat.dis-nn|year=2012}}</ref><br />
<br />
2012年6月，普渡大学的'''<font color="#ff8000">自旋电子学Spintronic</font>'''研究人员发表了一篇关于利用'''<font color="#ff8000">侧向自旋阀Lateral spin valves</font>'''和'''<font color="#ff8000">记忆电阻器Memristors</font>'''设计神经形态芯片的论文。他们认为，这种芯片结构的工作原理与神经元相似，因此可以用于大脑运行机制的复刻方法的测试。此外，这些芯片在能耗方面明显优于传统芯片。<ref name="Spin Devices Prop" /><br />
<br />
Research at [[HP Labs]] on Mott memristors has shown that while they can be non-[[Volatile memory|volatile]], the volatile behavior exhibited at temperatures significantly below the [[phase transition]] temperature can be exploited to fabricate a [[neuristor]],<ref name=":0" /> a biologically-inspired device that mimics behavior found in neurons.<ref name=":0">{{Cite journal | doi = 10.1038/nmat3510| pmid = 23241533| title = A scalable neuristor built with Mott memristors| journal = Nature Materials| volume = 12| issue = 2| pages = 114–7| year = 2012| last1 = Pickett | first1 = M. D. | last2 = Medeiros-Ribeiro | first2 = G. | last3 = Williams | first3 = R. S. | bibcode = 2013NatMa..12..114P| s2cid = 16271627| url = https://semanticscholar.org/paper/b6ba6f496ace2b947f111059663e76bb60e9efeb}}</ref> In September 2013, they presented models and simulations that show how the spiking behavior of these neuristors can be used to form the components required for a [[Turing machine]].<ref name=":14">{{cite journal|doi=10.1088/0957-4484/24/38/384002|title=Phase transitions enable computational universality in neuristor-based cellular automata|author1=Matthew D Pickett|author2=R Stanley Williams|name-list-style=amp|date=September 2013|publisher=IOP Publishing Ltd|journal=Nanotechnology|volume=24|issue=38|pmid=23999059|bibcode=2013Nanot..24L4002P|at=384002}}</ref><br />
<br />
'''<font color="#ff8000">惠普实验室HP labs</font>'''在莫特记忆电阻器上的研究表明，尽管它们可以是非'''<font color="#ff8000">易失性Volatile</font>'''的，但是在'''<font color="#ff8000">相变Phase transition</font>'''温度以下时表现出的易失性行为可以被用来制造'''<font color="#ff8000">类神经元电阻器Neuristor</font>'''(一种生物学启发的模仿神经元行为的硬件)<ref name=":0" />。2013年9月，他们通过模型和仿真展示了这些类神经元电阻器的脉冲行为如何产生'''<font color="#ff8000">图灵机Turing machine</font>'''的所需元素。<ref name=":14" /><br />
<br />
[[Neurogrid]], built by ''Brains in Silicon'' at [[Stanford University]],<ref name=":15">{{cite journal|last1=Boahen|first1=Kwabena|title=Neurogrid: A Mixed-Analog-Digital Multichip System for Large-Scale Neural Simulations|journal=Proceedings of the IEEE|date=24 April 2014|volume=102|issue=5|pages=699–716|doi=10.1109/JPROC.2014.2313565|s2cid=17176371}}</ref> is an example of hardware designed using neuromorphic engineering principles. The circuit board is composed of 16 custom-designed chips, referred to as NeuroCores. Each NeuroCore's analog circuitry is designed to emulate neural elements for 65536 neurons, maximizing energy efficiency. The emulated neurons are connected using digital circuitry designed to maximize spiking throughput.<ref name=":16">{{cite journal|doi=10.1038/503022a|pmid = 24201264|title = Neuroelectronics: Smart connections|journal = Nature|volume = 503|issue = 7474|pages = 22–4|year = 2013|last1 = Waldrop|first1 = M. Mitchell|bibcode = 2013Natur.503...22W|doi-access = free}}</ref><ref name=":17">{{cite journal|doi=10.1109/JPROC.2014.2313565|title = Neurogrid: A Mixed-Analog-Digital Multichip System for Large-Scale Neural Simulations|journal = Proceedings of the IEEE|volume = 102|issue = 5|pages = 699–716|year = 2014|last1 = Benjamin|first1 = Ben Varkey|last2 = Peiran Gao|last3 = McQuinn|first3 = Emmett|last4 = Choudhary|first4 = Swadesh|last5 = Chandrasekaran|first5 = Anand R.|last6 = Bussat|first6 = Jean-Marie|last7 = Alvarez-Icaza|first7 = Rodrigo|last8 = Arthur|first8 = John V.|last9 = Merolla|first9 = Paul A.|last10 = Boahen|first10 = Kwabena|s2cid = 17176371}}</ref><br />
<br />
'''<font color="#ff8000">神经栅格Neurogrid<ref name=":15" /></font>'''是由斯坦福大学Brains in Silicon公司研发的、使用神经形态工程原理设计的硬件。该电路板由16个定制设计的芯片组成(NeuroCores)。在设计中，每个NeuroCore芯片的模拟电路对65536个神经元的神经元素进行模拟，以最大限度地提高能量效率。模拟出的神经元通过设计的数字电路连接，以最大化脉冲吞吐量。<ref name=":16" /><ref name=":17" /><br />
<br />
A research project with implications for neuromorphic engineering is the [[Human Brain Project]] that is attempting to simulate a complete human brain in a supercomputer using biological data. It is made up of a group of researchers in neuroscience, medicine, and computing.<ref name=":18">{{cite web|title=Involved Organizations|url=http://www.humanbrainproject.eu/partners.html|access-date=22 February 2013|url-status=dead|archive-url=https://web.archive.org/web/20130302142627/http://www.humanbrainproject.eu/partners.html|archive-date=2 March 2013}}</ref> [[Henry Markram]], the project's co-director, has stated that the project proposes to establish a foundation to explore and understand the brain and its diseases, and to use that knowledge to build new computing technologies. The three primary goals of the project are to better understand how the pieces of the brain fit and work together, to understand how to objectively diagnose and treat brain diseases, and to use the understanding of the human brain to develop neuromorphic computers. That the simulation of a complete human brain will require a supercomputer a thousand times more powerful than today's encourages the current focus on neuromorphic computers.<ref name=":19">{{cite web|title=Human Brain Project|url=http://www.humanbrainproject.eu|access-date=22 February 2013}}</ref> $1.3 billion has been allocated to the project by The [[European Commission]].<ref name=":20">{{cite web|title=The Human Brain Project and Recruiting More Cyberwarriors|url=http://www.marketplace.org/topics/tech/human-brain-project-and-recruiting-more-cyberwarriors|access-date=22 February 2013|date=January 29, 2013}}</ref><br />
<br />
'''<font color="#ff8000">人类大脑计划Human Brain Project</font>'''对神经形态工程具有较大影响，其主要任务是尝试用生物数据在超级计算机中模拟完整的人脑。人类大脑计划由神经科学、医学和计算机科学背景的研究人员组成。<ref name=":18" />该项目的联合主管亨利•马克拉姆(Henry Markram)表示，人类大脑计划的目的是建立一个探索和了解脑科学和脑疾病知识的基础，并利用这些知识来构建更先进的计算机技术。这个项目的三个主要目标分别是: 更好地理解大脑的各个部分是如何相互配合协同工作的; 理解如何客观地诊断和治疗脑部疾病; 以及利用对人类大脑的理解来开发神经形态计算机。模拟一个完整的人类大脑需要一台比现在强大一千倍的超级计算机，这不断激发着对神经形态计算机领域的研究兴趣。<ref name=":19" />欧盟委员会已经向人类大脑计划拨款13亿美元。<ref name=":20" /><br />
<br />
Other research with implications for neuromorphic engineering involves the [[BRAIN Initiative]]<ref name="economist">[https://www.economist.com/news/science-and-technology/21582495-computers-will-help-people-understand-brains-better-and-understanding-brains Neuromorphic computing: The machine of a new soul], The Economist, 2013-08-03</ref> and the [[TrueNorth]] chip from [[IBM]].<ref name=":21">{{cite journal|last1=Modha|first1=Dharmendra|title=A million spiking-neuron integrated circuit with a scalable communication network and interface|journal=Science|date=Aug 2014|volume=345|issue=6197|pages=668–673|doi=10.1126/science.1254642|pmid=25104385|bibcode=2014Sci...345..668M|s2cid=12706847}}</ref> Neuromorphic devices have also been demonstrated using nanocrystals, nanowires, and conducting polymers.<ref name=":22">{{Cite web|url=http://jessamynfairfield.com/wp-content/uploads/2017/03/PWMar17Fairfield.pdf|title=Smarter Machines|last=Fairfield|first=Jessamyn|date=March 1, 2017}}</ref><br />
<br />
其他与神经形态工程有关的研究还包括'''<font color="#ff8000">脑计划BRAIN initiative</font>'''，<ref name="economist" />和IBM研发的'''<font color="#ff8000">TrueNorth</font>'''芯片。<ref name=":21" />使用纳米晶体、纳米线和导电聚合物也能够用于制造神经形态学硬件。<ref name=":22" /><br />
<br />
[[Intel]] unveiled its neuromorphic research chip, called “[[Intel Loihi|Loihi]]”, in October 2017. The chip uses an asynchronous [[spiking neural network]] (SNN) to implement adaptive self-modifying event-driven fine-grained parallel computations used to implement learning and inference with high efficiency.<ref name=":23">{{cite journal |last1=Davies |first1=Mike |title=Loihi: A Neuromorphic Manycore Processor with On-Chip Learning |journal=IEEE Micro |date=January 16, 2018 |volume=38 |issue=1 |pages=82–99 |display-authors=etal|doi=10.1109/MM.2018.112130359 |s2cid=3608458 }}</ref><ref name="Morris2017">{{cite web |last1=Morris |first1=John |title=Why Intel built a neuromorphic chip |url=https://www.zdnet.com/article/why-intel-built-a-neuromorphic-chip/ |website=ZDNet |access-date=17 August 2018 |language=en}}</ref><br />
<br />
2017年10月，英特尔发布了神经形态芯片'''<font color="#ff8000">Loihi</font>'''。该芯片采用异步'''<font color="#ff8000">脉冲神经网络Spiking neural network</font>'''实现了自适应、自修改、事件驱动的细粒度并行计算，实现了高效的学习和推理。<ref name=":23" /><ref name="Morris2017" /><br />
<br />
[[IMEC]], a Belgium-based nanoelectronics research center, demonstrated the world's first self-learning neuromorphic chip. The brain-inspired chip, based on OxRAM technology, has the capability of self-learning and has been demonstrated to have the ability to compose music.<ref name=":24">{{cite web |title=Imec demonstrates self-learning neuromorphic chip that composes music |url=https://www.imec-int.com/en/articles/imec-demonstrates-self-learning-neuromorphic-chip-that-composes-music |website=IMEC International |access-date=1 October 2019}}</ref> IMEC released the 30-second tune composed by the prototype. The chip was sequentially loaded with songs in the same time signature and style. The songs were old Belgian and French flute minuets, from which the chip learned the rules at play and then applied them.<ref name=":25">{{cite web|last1=Bourzac|first1=Katherine|title=A Neuromorphic Chip That Makes Music|url=https://spectrum.ieee.org/a-neuromorphic-chip-that-makes-music|url-status=live|access-date=1 October 2019|website=IEEE Spectrum|date=May 23, 2017}}</ref><br />
<br />
比利时的'''<font color="#ff8000">微电子研究中心IMEC</font>'''研发了世界上首个自学习神经形态芯片。这种基于'''<font color="#32CD32"> OxRAM(filamentary-oxide-based resistive memory technology)</font>''' 技术的大脑启发芯片具有自学习能力，并且已被证明具有创作音乐的能力。<ref name=":24" />IMEC发布了由芯片原型机谱写的30秒旋律。向芯片加载一系列特征、风格统一的歌曲（古代比利时和法国长笛小步舞曲），芯片从中学习相关规则并将其应用于创作。<ref name=":25" /><br />
<br />
[[Blue Brain Project|The Blue Brain Project]], led by Henry Markram, aims to build biologically detailed digital reconstructions and simulations of the mouse brain. The Blue Brain Project has created in silico models of rodent brains, while attempting to replicate as many details about its biology as possible. The supercomputer-based simulations offer new perspectives on understanding the structure and functions of the brain.<br />
<br />
由亨利·马克拉姆领导的'''<font color="#ff8000">蓝脑计划The Blue Brain Project</font>'''旨在建立小鼠大脑生理学细节的数字重建和模拟。蓝脑计划已经建立了啮齿动物大脑的电子模型，同时进行着尽可能多地复制其生理学细节的尝试。基于超级计算机的模拟为理解大脑的结构和功能提供了新的视角。<br />
<br />
The European Union funded a series of projects at the University of Heidelberg, which led to the development of <br />
[[BrainScaleS]] (brain-inspired multiscale computation in neuromorphic hybrid systems), a hybrid analog [[neuromorphic]] supercomputer located at Heidelberg University, Germany. It was developed as part of the [[Human Brain Project]] neuromorphic computing platform and is the complement to the [[SpiNNaker]] supercomputer (which is based on digital technology). The architecture used in BrainScaleS mimics biological neurons and their connections on a physical level; additionally, since the components are made of silicon, these model neurons operate on average 864 times (24 hours of real time is 100 seconds in the machine simulation) that of their biological counterparts.<ref name=":26">{{Cite web|date=2016-03-21|title=Beyond von Neumann, Neuromorphic Computing Steadily Advances|url=https://www.hpcwire.com/2016/03/21/lacking-breakthrough-neuromorphic-computing-steadily-advance/|access-date=2021-10-08|website=HPCwire|language=en-US}}</ref><br />
<br />
欧盟资助了海德堡大学一系列促进BrainScaleS(神经形态混合系统中受大脑启发的多尺度计算)发展的项目，这是一台位于德国海德堡大学的混合模拟'''<font color="#ff8000">神经形态Neuromorphic</font>'''超级计算机。它是作为人类大脑计划中神经形态计算平台的一部分而开发的，是'''<font color="#ff8000">SpiNNaker</font>'''超级计算机(基于数字技术)的补充。BrainScaleS中使用的体系架构模拟了生物神经元及其在物理层面上的连接；此外，由于这些组件是由硅制成的，这些模型神经元平均运行速度是生物神经元的864倍，这意味着在机器模拟中，24小时的实时时间仅为100秒。<ref name=":26" /><br />
<br />
===Neuromorphic sensors===<br />
===神经形态传感器===<br />
The concept of neuromorphic systems can be extended to sensors (not just to computation). An example of this applied to detecting [[light]] is the [[retinomorphic sensor]] or, when employed in an array, the [[event camera]].<br />
<br />
神经形态系统的概念可以扩展到传感器(而不仅仅是计算单元)。一个用于检测光线的例子是'''<font color="#ff8000">类视网膜传感器Retinomorphic sensor</font>'''，或者'''<font color="#ff8000">事件摄像机Event camera</font>阵列。'''<br />
<br />
<br />
==Ethical considerations ==<br />
==伦理问题==<br />
While the interdisciplinary concept of neuromorphic engineering is relatively new, many of the same ethical considerations apply to neuromorphic systems as apply to [[human-like machines]] and [[artificial intelligence]] in general. However, the fact that neuromorphic systems are designed to mimic a [[human brain]] gives rise to unique ethical questions surrounding their usage.<br />
<br />
虽然神经形态工程这个跨学科概念相对较新，但许多适用于'''<font color="#ff8000">类人机器Human-like machines</font>'''和人工智能的伦理考虑大体上也适用于神经形态系统。另一方面，神经形态系统是为了模仿人类大脑而设计，这一事实引起了一些针对性的独特伦理问题。<br />
<br />
However, the practical debate is that neuromorphic hardware as well as artificial "neural networks" are immensely simplified models of how the brain operates or processes information at a much lower [[complex system|complexity in terms of size and functional technology]] and a much more regular structure in terms of [[brain connectivity|connectivity]]. Comparing [[neuromorphic chip]]s to the brain is a very crude comparison similar to comparing a plane to a bird just because they both have wings and a tail. The fact is that neural cognitive systems are many orders of magnitude more [[Energy efficiency (physics)|energy-]] and compute-efficient than current state-of-the-art AI and neuromorphic engineering is an attempt to narrow this gap by inspiring from the brain's mechanism just like many engineering designs have [[bioengineering|bio-inspired features]].<br />
<br />
然而，这种争论的实际情况是，神经形态硬件和人工“神经网络”是对大脑运作或处理信息过程的极其简化的模型，在大小和功能技术方面的'''<font color="#ff8000">复杂性Complexity</font>'''相比而言比较低，在'''<font color="#ff8000">连接性Connectivity</font>'''方面也具有更加规则的结构。将'''<font color="#ff8000">神经形态芯片Neuromorphic chips</font>'''与大脑进行比较是一种非常粗糙的比较，类似于仅仅因为飞机有翅膀和尾巴就将它与鸟进行比较。事实上，当前最先进的人工智能在'''<font color="#ff8000">能耗效率Energy-efficiency</font>'''和计算效率方面距离人脑神经认知系统仍有较大差距，而神经形态工程只是一种通过从大脑机制中获得灵感来缩小这种差距的尝试，就像许多工程设计中都具有'''<font color="#ff8000">生物启发的特征Bio-inspired features</font>'''一样。<br />
<br />
===Democratic concerns===<br />
===公众担忧===<br />
Significant ethical limitations may be placed on neuromorphic engineering due to public perception.<ref name=":27">{{Cite report|url=https://ai100.stanford.edu/sites/g/files/sbiybj9861/f/ai_100_report_0831fnl.pdf|title=Artificial Intelligence and Life in 2030|author=2015 Study Panel|date=September 2016|work=One Hundred Year Study on Artificial Intelligence (AI100)|publisher=Stanford University}}</ref> Special [[Eurobarometer]] 382: Public Attitudes Towards Robots, a survey conducted by the European Commission, found that 60% of [[European Union]] citizens wanted a ban of robots in the care of children, the elderly, or the disabled. Furthermore, 34% were in favor of a ban on robots in education, 27% in healthcare, and 20% in leisure. The European Commission classifies these areas as notably “human.” The report cites increased public concern with robots that are able to mimic or replicate human functions. Neuromorphic engineering, by definition, is designed to replicate the function of the human brain.<ref name=":1">{{Cite web|url=http://ec.europa.eu/commfrontoffice/publicopinion/archives/ebs/ebs_382_en.pdf|title=Special Eurobarometer 382: Public Attitudes Towards Robots|last=European Commission|date=September 2012|website=European Commission}}</ref><br />
<br />
由于公众认知的相关忧虑，神经形态工程学可能会受到严重的伦理限制。<ref name=":27" />欧盟委员会进行的一项调查发现，60% 的欧盟公民希望禁止机器人参与照顾儿童、老人或残疾人的工作。此外，34% 的人支持禁止机器人用于教育，27% 的人支持禁止机器人用于医疗保健，20% 的人支持禁止机器人用于休闲。欧盟委员会将以上领域划入“人类”范畴。报告指出，公众越来越关注能够模仿或复制人类功能的机器人。神经形态工程，顾名思义，是为了复制人脑的功能而设计的。<ref name=":1" /><br />
<br />
The democratic concerns surrounding neuromorphic engineering are likely to become even more profound in the future. The European Commission found that EU citizens between the ages of 15 and 24 are more likely to think of robots as human-like (as opposed to instrument-like) than EU citizens over the age of 55. When presented an image of a robot that had been defined as human-like, 75% of EU citizens aged 15–24 said it corresponded with the idea they had of robots while only 57% of EU citizens over the age of 55 responded the same way. The human-like nature of neuromorphic systems, therefore, could place them in the categories of robots many EU citizens would like to see banned in the future.<ref name=":1" /><br />
<br />
围绕神经形态工程的公众担忧可能在未来变得更加深刻。欧盟委员会发现，相比于55岁以上的欧盟公民，15至24岁的欧盟公民更有可能认为机器人像人(而不是像仪器)。当看到一张“类人”机器人的图片时，年龄在15岁至24岁之间的欧盟公民中有75% 的人表示这符合他们对机器人的想法，而55岁以上的欧盟公民中只有57% 的人有同样的反应。因此，神经形态系统可能因为其类似人类的特性而被归入许多欧盟公民希望在未来禁止使用的机器人类别。<ref name=":1" /><br />
<br />
=== Personhood===<br />
=== 人格问题===<br />
As neuromorphic systems have become increasingly advanced, some scholars{{who|date=August 2021}} have advocated for granting [[personhood]] rights to these systems. If the brain is what grants humans their personhood, to what extent does a neuromorphic system have to mimic the human brain to be granted personhood rights? Critics of technology development in the [[Human Brain Project]], which aims to advance brain-inspired computing, have argued that advancement in neuromorphic computing could lead to machine consciousness or personhood.<ref name=":28">{{Cite journal|last=Aicardi|first=Christine|date=September 2018|title=Accompanying technology development in the Human Brain Project: From foresight to ethics management|journal=Futures|volume=102|pages=114–124|doi=10.1016/j.futures.2018.01.005|doi-access=free}}</ref> If these systems are to be treated as people, critics argue, then many tasks humans perform using neuromorphic systems, including the act of termination of neuromorphic systems, may be morally impermissible as these acts would violate the autonomy of the neuromorphic systems.<ref name=":29">{{Cite journal|last=Lim|first=Daniel|date=2014-06-01|title=Brain simulation and personhood: a concern with the Human Brain Project|journal=Ethics and Information Technology|language=en|volume=16|issue=2|pages=77–89|doi=10.1007/s10676-013-9330-5|s2cid=17415814|issn=1572-8439}}</ref><br />
<br />
随着神经形态系统的日益发展，一些学者主张赋予这些系统'''<font color="ff8000">人格Personhood</font>'''权。如果是大脑赋予了人类人格，那么在多大程度上模仿人类大脑的神经形态系统才能被赋予人格权利？“人类大脑计划”旨在推进以大脑为灵感的计算机技术发展，该计划的批评者认为，神经形态计算机技术的进步可能导致机器意识或人格的形成。<ref name=":28" />这些批评者认为，如果这些系统被当作人来对待，那么人类使用神经形态系统执行任务（包括终止神经形态系统）的行为，在道德上就可能是不被允许的，因为这些行为将违反神经形态系统的自主性。<ref name=":29" /><br />
<br />
<br />
==Dual use (military applications)==<br />
==军民两用技术（军事应用）==<br />
The [[Joint Artificial Intelligence Center]], a branch of the U.S. military, is a center dedicated to the procurement and implementation of AI software and neuromorphic hardware for combat use. Specific applications include smart headsets/goggles and robots. JAIC intends to rely heavily on neuromorphic technology to connect "every fighter every shooter" within a network of neuromorphic-enabled units.<br />
<br />
'''<font color="#ff8000">联合人工智能中心The Joint Artificial Intelligence Center</font>'''(JAIC),是美国军队的一个分支，专门从事采购和实施用于战斗的人工智能软件和神经形态硬件。具体应用包括智能耳机/护目镜和机器人。JAIC打算高度依赖神经形态技术，'''<font color="#32CD32">使用神经形态技术来连接神经形态单位网络中的“每个战士每个射手”</font>'''。<br />
<br />
<br />
==Legal considerations==<br />
==法律问题==<br />
Skeptics have argued that there is no way to apply the electronic personhood, the concept of personhood that would apply to neuromorphic technology, legally. In a letter signed by 285 experts in law, robotics, medicine, and ethics opposing a European Commission proposal to recognize “smart robots” as legal persons, the authors write, “A legal status for a robot can’t derive from the [[Natural person|Natural Person]] model, since the robot would then hold [[human rights]], such as the right to dignity, the right to its integrity, the right to remuneration or the right to citizenship, thus directly confronting the Human rights. This would be in contradiction with the [[Charter of Fundamental Rights of the European Union]] and the [[Convention for the Protection of Human Rights and Fundamental Freedoms]].”<ref name=":30">{{Cite web|url=http://www.robotics-openletter.eu/|title=Robotics Openletter {{!}} Open letter to the European Commission|language=fr-FR|access-date=2019-05-10}}</ref><br />
<br />
质疑者者认为，在法律上没有办法应用能够适用于神经形态技术的电子人格。在一封由285名法律、机器人技术、医学和伦理学专家的联名信中，作者们反对欧盟委员会提出的承认“智能机器人”为法人的提议。他们写道，“机器人的法律地位无法从自然人模型中推导出来，因为机器人将被赋予人权，如尊严权、完整权、报酬权或公民权，从而直接面临人权问题。这将有悖于《欧联基本权利宪章》和《欧洲保障人权和根本自由公约》”。<ref name=":30" /><br />
<br />
===Ownership and property rights===<br />
===所有权及财产权问题===<br />
There is significant legal debate around property rights and artificial intelligence. In ''Acohs Pty Ltd v. Ucorp Pty Ltd'', Justice Christopher Jessup of the [[Federal Court of Australia]] found that the [[source code]] for [[Material safety data sheets|Material Safety Data Sheets]] could not be [[Copyright law of Australia|copyrighted]] as it was generated by a [[software interface]] rather than a human author.<ref name=":31">{{Cite web|url=http://www.lavan.com.au/advice/intellectual_property_technology/copyright_in_source_code_and_digital_products|title=Copyright in source code and digital products|last=Lavan|website=Lavan|language=en|access-date=2019-05-10}}</ref> The same question may apply to neuromorphic systems: if a neuromorphic system successfully mimics a human brain and produces a piece of original work, who, if anyone, should be able to claim ownership of the work?<ref name=":32">{{cite journal |last1=Eshraghian|first1=Jason K. |title=Human Ownership of Artificial Creativity |journal=Nature Machine Intelligence |date=9 March 2020 |volume=2 |pages=157–160 |doi=10.1038/s42256-020-0161-x}}</ref><br />
<br />
法律界围绕财产权和人工智能有着重大争论。在Acohs Pty有限公司诉Ucorp Pty有限公司一案中，澳大利亚联邦法院的克里斯托弗·杰瑟普法官发现，版权保护不适用于材料安全数据表的源代码，因为它是由软件生成而非人类工作者生成的。<ref name=":31" />同样的问题可能也适用于神经形态系统: 如果一个神经形态系统成功地模仿了人类的大脑，并产生了一部原创作品，那么该如何确认这部作品的所有权归属？<ref name=":32" /><br />
<br />
<br />
==Neuromemristive systems==<br />
==神经记忆电阻系统 ==<br />
Neuromemristive systems are a subclass of neuromorphic computing systems that focus on the use of [[memristors]] to implement [[neuroplasticity]]. While neuromorphic engineering focuses on mimicking biological behavior, neuromemristive systems focus on abstraction.<ref>{{Cite web|url=https://digitalops.sandia.gov/Mediasite/Play/a10cf6ceb55d47608bb8326dd00e46611d|title=002.08 N.I.C.E. Workshop 2014: Towards Intelligent Computing with Neuromemristive Circuits and Systems - Feb. 2014|website=digitalops.sandia.gov|access-date=2019-08-26}}</ref> For example, a neuromemristive system may replace the details of a [[Cerebral cortex|cortical]] microcircuit's behavior with an abstract neural network model.<ref>C. Merkel and D. Kudithipudi, "Neuromemristive extreme learning machines for pattern classification," ISVLSI, 2014.</ref><br />
<br />
Neuromemristive systems are a subclass of neuromorphic computing systems that focus on the use of memristors to implement neuroplasticity. While neuromorphic engineering focuses on mimicking biological behavior, neuromemristive systems focus on abstraction. For example, a neuromemristive system may replace the details of a cortical microcircuit's behavior with an abstract neural network model.C. Merkel and D. Kudithipudi, "Neuromemristive extreme learning machines for pattern classification," ISVLSI, 2014.<br />
<br />
神经记忆电阻系统是神经形态计算系统的一个亚类，主要研究利用记忆电阻器实现神经可塑性。神经形态工程的重点是模拟生物行为，而神经记忆电阻系统的重点是提取。例如，一个神经记忆系统可以用一个抽象的神经网络模型取代皮层微电路的行为细节。默克尔和 d. Kudithipudi，“用于模式分类的神经记忆极端学习机器”，ISVLSI，2014。<br />
<br />
There exist several neuron inspired threshold logic functions<ref name="Maan 1–13" /> implemented with memristors that have applications in high level [[pattern recognition]] applications. Some of the applications reported recently include [[speech recognition]],<ref>{{Cite journal|title = Memristor pattern recogniser: isolated speech word recognition|journal = Electronics Letters|pages = 1370–1372|volume = 51|issue = 17|doi = 10.1049/el.2015.1428|first1 = A.K.|last1 = Maan|first2 = A.P.|last2 = James|first3 = S.|last3 = Dimitrijev|year = 2015|bibcode = 2015ElL....51.1370M|hdl = 10072/140989|s2cid = 61454815|url = https://semanticscholar.org/paper/48d3ab11ec6e213b62f11eedcfb7b7febb058674|hdl-access = free}}</ref> [[face recognition]]<ref>{{Cite journal|title = Memristive Threshold Logic Face Recognition|journal = Procedia Computer Science|date = 2014-01-01|pages = 98–103|volume = 41|series = 5th Annual International Conference on Biologically Inspired Cognitive Architectures, 2014 BICA|doi = 10.1016/j.procs.2014.11.090|first1 = Akshay Kumar|last1 = Maan|first2 = Dinesh S.|last2 = Kumar|first3 = Alex Pappachen|last3 = James|doi-access = free}}</ref> and [[object recognition]].<ref>{{Cite journal|title = Memristive Threshold Logic Circuit Design of Fast Moving Object Detection|journal = IEEE Transactions on Very Large Scale Integration (VLSI) Systems|date = 2015-10-01|issn = 1063-8210|pages = 2337–2341|volume = 23|issue = 10|doi = 10.1109/TVLSI.2014.2359801|first1 = A.K.|last1 = Maan|first2 = D.S.|last2 = Kumar|first3 = S.|last3 = Sugathan|first4 = A.P.|last4 = James|arxiv = 1410.1267|s2cid = 9647290}}</ref> They also find applications in replacing conventional digital logic gates.<ref>{{Cite journal|title = Resistive Threshold Logic|journal = IEEE Transactions on Very Large Scale Integration (VLSI) Systems|date = 2014-01-01|issn = 1063-8210|pages = 190–195|volume = 22|issue = 1|doi = 10.1109/TVLSI.2012.2232946|first1 = A.P.|last1 = James|first2 = L.R.V.J.|last2 = Francis|first3 = D.S.|last3 = Kumar|arxiv = 1308.0090|s2cid = 7357110}}</ref><ref>{{Cite journal|title = Threshold Logic Computing: Memristive-CMOS Circuits for Fast Fourier Transform and Vedic Multiplication|journal = IEEE Transactions on Very Large Scale Integration (VLSI) Systems|date = 2015-11-01|issn = 1063-8210|pages = 2690–2694|volume = 23|issue = 11|doi = 10.1109/TVLSI.2014.2371857|first1 = A.P.|last1 = James|first2 = D.S.|last2 = Kumar|first3 = A.|last3 = Ajayan|arxiv = 1411.5255|s2cid = 6076956}}</ref><br />
<br />
There exist several neuron inspired threshold logic functions implemented with memristors that have applications in high level pattern recognition applications. Some of the applications reported recently include speech recognition, face recognition and object recognition. They also find applications in replacing conventional digital logic gates.<br />
<br />
利用记忆电阻器实现的神经元启发阈值逻辑函数在高级模式识别中有着广泛的应用。最近报道的一些应用包括语音识别、人脸识别和物体识别。它们还可以用来取代传统的数字逻辑门。<br />
<br />
For ideal passive memristive circuits there is an exact equation (Caravelli-Traversa-[[Di Ventra]] equation) for the internal memory of the circuit:<ref>{{cite journal |last=Caravelli |display-authors=etal|arxiv=1608.08651 |title=The complex dynamics of memristive circuits: analytical results and universal slow relaxation |year=2017 |doi=10.1103/PhysRevE.95.022140 |pmid= 28297937 |volume=95 |issue= 2 |pages= 022140 |journal=Physical Review E|bibcode=2017PhRvE..95b2140C |s2cid=6758362}}</ref><br />
<br />
For ideal passive memristive circuits there is an exact equation (Caravelli-Traversa-Di Ventra equation) for the internal memory of the circuit:<br />
<br />
对于理想的无源记忆电路，电路的内部记忆有一个精确的方程(Caravelli-Traversa-Di Ventra 方程) :<br />
<br />
:<math> \frac{d}{dt} \vec{W} = \alpha \vec{W}-\frac{1}{\beta} (I+\xi \Omega W)^{-1} \Omega \vec S </math><br />
<br />
:\frac{d}{dt} \vec{W} = \alpha \vec{W}-\frac{1}{\beta} (I+\xi \Omega W)^{-1} \Omega \vec S<br />
<br />
: \frac{d}{dt} \vec{W} = \alpha \vec{W}-\frac{1}{\beta} (I+\xi \Omega W)^{-1} \Omega \vec S<br />
<br />
as a function of the properties of the physical memristive network and the external sources. In the equation above, <math>\alpha</math> is the "forgetting" time scale constant, <math> \xi=r-1</math> and <math>r=\frac{R_\text{off}}{R_\text{on}}</math> is the ratio of ''off'' and ''on'' values of the limit resistances of the memristors, <math> \vec S </math> is the vector of the sources of the circuit and <math>\Omega</math> is a projector on the fundamental loops of the circuit. The constant <math>\beta</math> has the dimension of a voltage and is associated to the properties of the [[memristor]]; its physical origin is the charge mobility in the conductor. The diagonal matrix and vector <math>W=\operatorname{diag}(\vec W)</math> and <math>\vec W</math> respectively, are instead the internal value of the memristors, with values between 0 and 1. This equation thus requires adding extra constraints on the memory values in order to be reliable.<br />
<br />
<nowiki>as a function of the properties of the physical memristive network and the external sources. In the equation above, \alpha is the "forgetting" time scale constant, \xi=r-1 and r=\frac{R_\text{off}}{R_\text{on}} is the ratio of off and on values of the limit resistances of the memristors, \vec S is the vector of the sources of the circuit and \Omega is a projector on the fundamental loops of the circuit. The constant \beta has the dimension of a voltage and is associated to the properties of the memristor; its physical origin is the charge mobility in the conductor. The diagonal matrix and vector W=\operatorname{diag}(\vec W) and \vec W respectively, are instead the internal value of the memristors, with values between 0 and 1. This equation thus requires adding extra constraints on the memory values in order to be reliable.</nowiki><br />
<br />
<nowiki>作为物理记忆网络和外部源的性质的函数。在上述方程中，α 是“遗忘”时间尺度常数，xi = r-1，r = frac { r _ text { off }{ on }{ r _ text { on }}是记忆电阻器极限电阻的开关和开关值之比，vec s 是电路源的矢量，Omega 是电路基本环路的投影仪。常数 β 具有电压的尺寸，与记忆电阻器的特性有关; 它的物理起源是导体中的电荷迁移率。对角矩阵和向量 w = 操作者名{ diag }(vec w)和 vec w 分别是记忆电阻器的内值，值在0到1之间。因此，这个等式需要在内存值上添加额外的约束，以保证可靠性。</nowiki><br />
<br />
==See also ==<br />
== 相关词条==<br />
{{Columns-list|colwidth=18em|<br />
* [[AI accelerator (computer hardware)|AI accelerator]]<br />
* [[Artificial brain]]<br />
* [[Biomorphic]]<br />
* [[Cognitive computer]]<br />
* [[Computation and Neural Systems]]<br />
* [[Differentiable programming]]<br />
* [[Event camera]]<br />
* [[Neurorobotics]]<br />
* [[Optical flow sensor]]<br />
* [[Physical neural network]]<br />
* [[SpiNNaker]]<br />
* [[SyNAPSE]]<br />
* [[Retinomorphic sensor]]<br />
* [[Vision chip]]<br />
* [[Vision processing unit]]<br />
* [[Zeroth (software)]]<br />
* [[Hardware for artificial intelligence]]<br />
}}<br />
<br />
{{Portal bar|Electronics}}<br />
<br />
==References==<br />
{{Reflist|40em}}<br />
<br />
==External links==<br />
<!--======================== {{No more links}} ============================<br />
| PLEASE BE CAUTIOUS IN ADDING MORE LINKS TO THIS ARTICLE. Wikipedia |<br />
| is not a collection of links nor should it be used for advertising. |<br />
| |<br />
| Excessive or inappropriate links WILL BE DELETED. |<br />
| See [[Wikipedia:External links]] & [[Wikipedia:Spam]] for details. |<br />
| |<br />
| If there are already plentiful links, please propose additions or |<br />
| replacements on this article's discussion page, or submit your link |<br />
| to the relevant category at the Open Directory Project (dmoz.org) |<br />
| and link back to that category using the {{dmoz}} template. |<br />
======================={{No more links}}=============================--><br />
{{Commons category|Neuromorphic Engineering}}<br />
*[https://web.archive.org/web/20150727034331/http://ine-web.org/workshops/workshops-overview Telluride Neuromorphic Engineering Workshop]<br />
*[https://archive.today/20130115190057/http://capocaccia.ethz.ch/ CapoCaccia Cognitive Neuromorphic Engineering Workshop]<br />
*[https://web.archive.org/web/20190716132350/http://www.ine-web.org/ Institute of Neuromorphic Engineering]<br />
*[http://www.ine-news.org/ INE news site].<br />
*[http://www.frontiersin.org/neuromorphic_engineering Frontiers in Neuromorphic Engineering Journal]<br />
*[http://www.cns.caltech.edu/ Computation and Neural Systems] department at the [[California Institute of Technology]].<br />
*[http://www.humanbrainproject.eu/ Human Brain Project official site]<br />
*[https://www.the-scientist.com/features/building-a-silicon-brain-65738 Building a Silicon Brain:] Computer chips based on biological neurons may help simulate larger and more-complex brain models. May 1, 2019. SANDEEP RAVINDRAN<br />
<br />
<br />
*Telluride Neuromorphic Engineering Workshop<br />
*CapoCaccia Cognitive Neuromorphic Engineering Workshop<br />
*Institute of Neuromorphic Engineering<br />
*INE news site.<br />
*Frontiers in Neuromorphic Engineering Journal<br />
*Computation and Neural Systems department at the California Institute of Technology.<br />
*Human Brain Project official site<br />
* Building a Silicon Brain: Computer chips based on biological neurons may help simulate larger and more-complex brain models. May 1, 2019. SANDEEP RAVINDRAN<br />
<br />
=<nowiki>外部链接</nowiki>= <br />
*碲化物神经形态工程工作室<br />
*CapoCaccia 认知神经形态工程工作室<br />
*神经形态工程研究所<br />
*INE 新闻站点。<br />
*《神经形态工程学前沿》<br />
*加州理工学院计算与神经系统系。<br />
*人脑项目官方网站<br />
*打造硅脑: 基于生物神经元的计算机芯片可能有助于模拟更大、更复杂的大脑模型。2019年5月1日。SANDEEP RAVINDRAN<br />
<br />
{{Differentiable computing}}<br />
{{Authority control}}<br />
<br />
[[Category:Electrical engineering]]<br />
[[Category:Neuroscience]]<br />
[[分类:AI accelerators|*]]<br />
[[Category:Artificial intelligence]]<br />
[[Category:Robotics]]<br />
<br />
Category:Electrical engineering<br />
Category:Neuroscience<br />
*<br />
Category:Artificial intelligence<br />
Category:Robotics<br />
<br />
类别: 电气工程类别: 神经科学 * 类别: 人工智能类别: 机器人<br />
<br />
<noinclude><br />
<br />
<small>This page was moved from [[wikipedia:en:Neuromorphic engineering]]. Its edit history can be viewed at [[神经形态计算/edithistory]]</small></noinclude><br />
<br />
[[Category:待整理页面]]<br />
<br />
<references /></div>Shenky20https://wiki.swarma.org/index.php?title=%E7%A5%9E%E7%BB%8F%E5%BD%A2%E6%80%81%E8%AE%A1%E7%AE%97&diff=30747神经形态计算2022-04-28T10:49:51Z<p>Shenky20：/* 军民两用技术（军事应用） */</p>
<hr />
<div><br />
此词条由神经动力学模型读书会词条梳理志愿者Shenky20翻译审校，未经专家审校，带来阅读不便，请见谅。<br />
<br />
{{Use American English|date = January 2019}}<br />
{{Use mdy dates|date = January 2019}}<br />
<br />
'''Neuromorphic engineering''', also known as '''neuromorphic computing''',<ref name=":3">{{Cite journal | doi = 10.1145/2601069| title = Neuromorphic computing gets ready for the (really) big time| journal = [[Communications of the ACM]]| volume = 57| issue = 6| year = 2014| last1 = Monroe | first1 = D. | pages = 13–15| s2cid = 20051102}}</ref><ref name=":4">{{Cite journal | doi = 10.1088/0957-4484/21/17/175202| pmid = 20368686| title = Nanotube devices based crossbar architecture: Toward neuromorphic computing| journal = Nanotechnology| volume = 21| issue = 17| pages = 175202| year = 2010| last1 = Zhao | first1 = W. S. | last2 = Agnus | first2 = G. | last3 = Derycke | first3 = V. | last4 = Filoramo | first4 = A. | last5 = Bourgoin | first5 = J. -P. | last6 = Gamrat | first6 = C. | bibcode = 2010Nanot..21q5202Z| url = https://zenodo.org/record/3428659}}</ref><ref name="humanbrainproject">{{YouTube|id=6RoiZ90mGfw|title=The Human Brain Project SP 9: Neuromorphic Computing Platform}}</ref> is the use of [[very-large-scale integration]] (VLSI) systems containing electronic [[analog circuit]]s to mimic neuro-biological architectures present in the nervous system. A neuromorphic computer/chip is any device that uses physical artificial neurons (made from silicon) to do computations.<ref name=":5">{{cite journal|last1=Mead|first1=Carver|title=Neuromorphic electronic systems|journal=Proceedings of the IEEE|date=1990|volume=78|issue=10|pages=1629–1636|doi=10.1109/5.58356|url=https://authors.library.caltech.edu/53090/1/00058356.pdf}}</ref><ref name=":2" /> In recent times, the term ''neuromorphic'' has been used to describe analog, digital, mixed-mode analog/digital VLSI, and software systems that implement models of [[neural system]]s (for [[perception]], [[motor control]], or [[multisensory integration]]). The implementation of neuromorphic computing on the hardware level can be realized by oxide-based [[memristor]]s,<ref name="Maan 1–13">{{Cite journal|last1=Maan|first1=A. K.|last2=Jayadevi|first2=D. A.|last3=James|first3=A. P.|date=2016-01-01|title=A Survey of Memristive Threshold Logic Circuits|journal=IEEE Transactions on Neural Networks and Learning Systems|volume=PP|issue=99|pages=1734–1746|doi=10.1109/TNNLS.2016.2547842|pmid=27164608|issn=2162-237X|arxiv=1604.07121|bibcode=2016arXiv160407121M|s2cid=1798273}}</ref> spintronic memories, threshold switches, and [[transistor]]s.<ref name=":6">{{Cite journal|title = Mott Memory and Neuromorphic Devices|journal = Proceedings of the IEEE|date = 2015-08-01|issn = 0018-9219|pages = 1289–1310|volume = 103|issue = 8|doi = 10.1109/JPROC.2015.2431914|first1 = You|last1 = Zhou|first2 = S.|last2 = Ramanathan|s2cid = 11347598|url = https://zenodo.org/record/895565}}</ref><ref name=":2">{{Cite document|title=Neuromorphic Circuits With Neural Modulation Enhancing the Information Content of Neural Signaling {{!}} International Conference on Neuromorphic Systems 2020|language=EN|doi=10.1145/3407197.3407204|s2cid=220794387}}</ref> Training software-based neuromorphic systems of spiking neural networks can be achieved using error backpropagation, e.g., using [[Python (programming language)|Python]] based frameworks such as snnTorch,<ref name=":7">{{cite journal |last1=Eshraghian|first1=Jason K.|last2=Ward|first2=Max|last3=Neftci |first3=Emre|last4=Wang|first4=Xinxin|last5=Lenz|first5=Gregor|last6=Dwivedi|first6=Girish|last7=Bennamoun|first7=Mohammed|last8=Jeong|first8=Doo Seok|last9=Lu|first9=Wei D.|title=Training Spiking Neural Networks Using Lessons from Deep Learning |date=1 October 2021 |arxiv=2109.12894 }}</ref> or using canonical learning rules from the biological learning literature, e.g., using BindsNet.<ref name=":8">{{Cite web | url=https://github.com/Hananel-Hazan/bindsnet | title=Hananel-Hazan/bindsnet: Simulation of spiking neural networks (SNNs) using PyTorch.| date=31 March 2020}}</ref><br />
<br />
<br />
'''<font color="#ff8000">神经形态工程Neuromorphic engineering</font>'''(也称为'''<font color="#ff8000">神经形态计算Neuromorphic computing</font>''')<ref name=":3" /><ref name=":4" /><ref name="humanbrainproject" />是指使用包含电子'''<font color="#ff8000">模拟电路Analog circuit</font>'''的'''<font color="#ff8000">超大规模集成电路Very-large-scale integration</font>'''系统来模拟神经系统中生理结构的研究方法。神经形态计算机或神经形态芯片包括任何使用由硅制成的人造神经元进行计算的设备。<ref name=":5" /><ref name=":2" />近年来，神经形态学（neuromorphic）这个术语被用来描述能够实现'''<font color="#ff8000">神经系统Neural system</font>'''模型功能(如'''<font color="#ff8000">感知Perception</font>'''、'''<font color="#ff8000">运动控制Motor control</font>'''，'''<font color="#ff8000">多感官整合Multisensory integration</font>'''等)的模拟、数字、模拟/数字混合模式超大规模集成电路和软件系统。神经形态计算的硬件实现可以通过基于氧化物的'''<font color="#ff8000">记忆电阻器Memristor</font>'''、<ref name="Maan 1–13" />自旋电子存储器、阈值开关和'''<font color="#ff8000">晶体管Transistor</font>'''来实现。<ref name=":6" /><ref name=":2" />对基于软件的脉冲神经网络系统的训练可以通过误差反向传播机制来实现，例如，使用snnTorch等基于Python的框架，<ref name=":7" />或使用BindsNet等典型的受生物启发的学习模式。<ref name=":8" /><br />
<br />
A key aspect of neuromorphic engineering is understanding how the morphology of individual neurons, circuits, applications, and overall architectures creates desirable computations, affects how information is represented, influences robustness to damage, incorporates learning and development, adapts to local change (plasticity), and facilitates evolutionary change.<br />
<br />
<br />
神经形态工程领域的一个关键问题，就是理解'''<font color="#32CD32">单个神经元形态、神经回路、应用和整体结构</font>'''如何产生理想的计算，如何影响信息的表示和对破坏的鲁棒性，如何整合学习和发展，如何适应局部变化(可塑性) 并促进逐渐发展的变化。<br />
<br />
Neuromorphic engineering is an interdisciplinary subject that takes inspiration from [[biology]], [[physics]], [[mathematics]], [[computer science]], and [[electronic engineering]]<ref name=":2" /> to design artificial neural systems, such as [[Machine vision|vision systems]], head-eye systems, auditory processors, and autonomous robots, whose physical architecture and design principles are based on those of biological nervous systems.<ref name=":9">{{Cite journal | doi = 10.1155/2012/705483| title = Qualitative Functional Decomposition Analysis of Evolved Neuromorphic Flight Controllers| journal = Applied Computational Intelligence and Soft Computing| volume = 2012| pages = 1–21| year = 2012| last1 = Boddhu | first1 = S. K. | last2 = Gallagher | first2 = J. C. | doi-access = free}}</ref> It was developed by [[Carver Mead]]<ref name=":10">{{cite web|title=carver mead website|last1=Mead|first1=Carver|url=http://www.carvermead.caltech.edu/index.html|website=carvermead}}</ref> in the late 1980s.<br />
<br />
<br />
神经形态工程是以'''<font color="#ff8000">生物学Biology</font>'''、'''<font color="#ff8000">物理学Physics</font>'''、'''<font color="#ff8000">数学Mathematics</font>'''、'''<font color="#ff8000">计算机科学Computer science</font>'''和'''<font color="#ff8000">电子工程Electronic engineering<ref name=":2" /></font>'''等学科为基础，设计人工神经系统（如'''<font color="#ff8000">视觉系统Vision systems</font>'''、头眼系统、听觉处理器以及物理结构和设计原则都受启发于生物神经系统的自主机器人）的一门交叉学科。<ref name=":9" />20世纪80年代后期，卡弗·米德极大地推动了神经形态工程领域的发展。<ref name=":10" /><br />
<br />
==Neurological inspiration==<br />
==神经生物学启发==<br />
Neuromorphic engineering is set apart by the inspiration it takes from what we know about the structure and operations of the [[brain]]. Neuromorphic engineering translates what we know about the brain's function into computer systems. Work has mostly focused on replicating the analog nature of [[biological computation]] and the role of [[neuron]]s in [[cognition]].<br />
<br />
<br />
神经形态工程形成的灵感来源于现已知的'''<font color="#ff8000">大脑Brain</font>'''结构及运作机制知识，它将我们对大脑功能的了解用于研究和优化计算机系统。该领域工作主要集中于对'''<font color="#ff8000">生物计算Biological computation</font>'''模拟特性和'''<font color="#ff8000">神经元Neuron</font>'''在'''<font color="#ff8000">认知Cognition</font>'''中所发挥作用的复刻。<br />
<br />
The biological processes of neurons and their [[synapse]]s are dauntingly complex, and thus very difficult to artificially simulate. A key feature of biological brains is that all of the processing in neurons use analog chemical signals. This makes it hard to replicate brains in computers because the current generation of computers is completely digital. However, the characteristics of these parts can be abstracted into mathematical functions that closely capture the essence of the neuron's operations.<br />
<br />
神经元及其'''<font color="#ff8000">突触Synapse</font>'''的生理学过程极其复杂，难以进行人工模拟。神经元在所有处理过程中都使用化学模拟信号，这是大脑的一个关键生理学特征。这个特征大大增加了在计算机中复制大脑的难度，因为目前的计算机是完全数字化的。然而，部分特征可以抽象为数学函数，这些函数能够紧密捕捉神经元运行的本质。<br />
<br />
The goal of neuromorphic computing is not to perfectly mimic the brain and all of its functions, but instead to extract what is known of its structure and operations to be used in a practical computing system. No neuromorphic system will claim nor attempt to reproduce every element of neurons and synapses, but all adhere to the idea that computation is highly [[distributed processing|distributed]] throughout a series of small computing elements analogous to a neuron. While this sentiment is standard, researchers chase this goal with different methods.<ref name=":11">{{Cite journal | doi = 10.1088/1741-2560/13/5/051001| title = Large-scale neuromorphic computing systems| journal = Journal of Neural Engineering| volume = 13| pages = 1–15| year = 2016| last1 = Furber | first1 = Steve| issue = 5| pmid = 27529195| bibcode = 2016JNEng..13e1001F| doi-access = free}}</ref><br />
<br />
神经形态计算的目标不是完美地模拟大脑及其所有功能，而是利用已知的大脑结构和运转机制来研发或优化实际的计算系统。任何神经形态学系统都不会声称或试图复制神经元和突触中的每一个元素，但所有人都一致认可将计算高度'''<font color="#ff8000">分散Distribute</font>'''于一系列类似于神经元的小型计算元素的理念。研究人员用不同的方法来追求这一普遍目标。<ref name=":11" /><br />
<br />
<br />
==Examples==<br />
==代表性成果==<br />
As early as 2006, researchers at [[Georgia Tech]] published a field programmable neural array.<ref name=":12">{{Cite book|title = A field programmable neural array|last1 = Farquhar|first1 = Ethan|date = May 2006|journal = IEEE International Symposium on Circuits and Systems|pages = 4114–4117|last2 = Hasler|first2 = Paul.|doi = 10.1109/ISCAS.2006.1693534|isbn = 978-0-7803-9389-9|s2cid = 206966013}}</ref> This chip was the first in a line of increasingly complex arrays of floating gate transistors that allowed programmability of charge on the gates of [[MOSFET]]s to model the channel-ion characteristics of neurons in the brain and was one of the first cases of a silicon programmable array of neurons.<br />
<br />
早在2006年，佐治亚理工学院的研究人员就研发出了一种现场可编程神经阵列。<ref name=":12" />在此之后，一系列越来越复杂的浮栅晶体管阵列被成功研发出来，这些晶体管阵列可以通过在'''<font color="#ff8000">金属-氧化物半导体效应晶体管MOSFET</font>'''的栅极上编程来模拟大脑中神经元的离子通道特性，这也是以硅为主要材料的可编程神经元阵列的最早成功案例之一。<br />
<br />
In November 2011, a group of [[MIT]] researchers created a computer chip that mimics the analog, ion-based communication in a synapse between two neurons using 400 transistors and standard [[CMOS]] manufacturing techniques.<ref name=":13">{{cite web|title=MIT creates "brain chip"|url=http://www.extremetech.com/extreme/105067-mit-creates-brain-chip|access-date=4 December 2012}}</ref><ref name="Neuromorphic silicon paper">{{cite journal|title=Neuromorphic silicon neurons and large-scale neural networks: challenges and opportunities|doi=10.3389/fnins.2011.00108|pmid=21991244|pmc=3181466|volume=5|pages=108|journal=Frontiers in Neuroscience|year=2011|last1=Poon|first1=Chi-Sang|last2=Zhou|first2=Kuan|doi-access=free}}</ref><br />
<br />
2011年11月，麻省理工学院的一组研究人员研发出一种计算机芯片，该芯片上使用标准的'''<font color="#ff8000">互补金属氧化物半导体CMOS</font>'''制造技术集成了400个晶体管，能够模拟神经元间突触中基于离子的通讯。<ref name=":13" /><ref name="Neuromorphic silicon paper" /><br />
<br />
In June 2012, [[spintronic]] researchers at [[Purdue University]] presented a paper on the design of a neuromorphic chip using [[Spin valve|lateral spin valve]]s and [[memristor]]s. They argue that the architecture works similarly to neurons and can therefore be used to test methods of reproducing the brain's processing. In addition, these chips are significantly more energy-efficient than conventional ones.<ref name="Spin Devices Prop">{{Cite arXiv|title=Proposal For Neuromorphic Hardware Using Spin Devices|eprint=1206.3227|last1=Sharad|first1=Mrigank|last2=Augustine|first2=Charles|last3=Panagopoulos|first3=Georgios|last4=Roy|first4=Kaushik|class=cond-mat.dis-nn|year=2012}}</ref><br />
<br />
2012年6月，普渡大学的'''<font color="#ff8000">自旋电子学Spintronic</font>'''研究人员发表了一篇关于利用'''<font color="#ff8000">侧向自旋阀Lateral spin valves</font>'''和'''<font color="#ff8000">记忆电阻器Memristors</font>'''设计神经形态芯片的论文。他们认为，这种芯片结构的工作原理与神经元相似，因此可以用于大脑运行机制的复刻方法的测试。此外，这些芯片在能耗方面明显优于传统芯片。<ref name="Spin Devices Prop" /><br />
<br />
Research at [[HP Labs]] on Mott memristors has shown that while they can be non-[[Volatile memory|volatile]], the volatile behavior exhibited at temperatures significantly below the [[phase transition]] temperature can be exploited to fabricate a [[neuristor]],<ref name=":0" /> a biologically-inspired device that mimics behavior found in neurons.<ref name=":0">{{Cite journal | doi = 10.1038/nmat3510| pmid = 23241533| title = A scalable neuristor built with Mott memristors| journal = Nature Materials| volume = 12| issue = 2| pages = 114–7| year = 2012| last1 = Pickett | first1 = M. D. | last2 = Medeiros-Ribeiro | first2 = G. | last3 = Williams | first3 = R. S. | bibcode = 2013NatMa..12..114P| s2cid = 16271627| url = https://semanticscholar.org/paper/b6ba6f496ace2b947f111059663e76bb60e9efeb}}</ref> In September 2013, they presented models and simulations that show how the spiking behavior of these neuristors can be used to form the components required for a [[Turing machine]].<ref name=":14">{{cite journal|doi=10.1088/0957-4484/24/38/384002|title=Phase transitions enable computational universality in neuristor-based cellular automata|author1=Matthew D Pickett|author2=R Stanley Williams|name-list-style=amp|date=September 2013|publisher=IOP Publishing Ltd|journal=Nanotechnology|volume=24|issue=38|pmid=23999059|bibcode=2013Nanot..24L4002P|at=384002}}</ref><br />
<br />
'''<font color="#ff8000">惠普实验室HP labs</font>'''在莫特记忆电阻器上的研究表明，尽管它们可以是非'''<font color="#ff8000">易失性Volatile</font>'''的，但是在'''<font color="#ff8000">相变Phase transition</font>'''温度以下时表现出的易失性行为可以被用来制造'''<font color="#ff8000">类神经元电阻器Neuristor</font>'''(一种生物学启发的模仿神经元行为的硬件)<ref name=":0" />。2013年9月，他们通过模型和仿真展示了这些类神经元电阻器的脉冲行为如何产生'''<font color="#ff8000">图灵机Turing machine</font>'''的所需元素。<ref name=":14" /><br />
<br />
[[Neurogrid]], built by ''Brains in Silicon'' at [[Stanford University]],<ref name=":15">{{cite journal|last1=Boahen|first1=Kwabena|title=Neurogrid: A Mixed-Analog-Digital Multichip System for Large-Scale Neural Simulations|journal=Proceedings of the IEEE|date=24 April 2014|volume=102|issue=5|pages=699–716|doi=10.1109/JPROC.2014.2313565|s2cid=17176371}}</ref> is an example of hardware designed using neuromorphic engineering principles. The circuit board is composed of 16 custom-designed chips, referred to as NeuroCores. Each NeuroCore's analog circuitry is designed to emulate neural elements for 65536 neurons, maximizing energy efficiency. The emulated neurons are connected using digital circuitry designed to maximize spiking throughput.<ref name=":16">{{cite journal|doi=10.1038/503022a|pmid = 24201264|title = Neuroelectronics: Smart connections|journal = Nature|volume = 503|issue = 7474|pages = 22–4|year = 2013|last1 = Waldrop|first1 = M. Mitchell|bibcode = 2013Natur.503...22W|doi-access = free}}</ref><ref name=":17">{{cite journal|doi=10.1109/JPROC.2014.2313565|title = Neurogrid: A Mixed-Analog-Digital Multichip System for Large-Scale Neural Simulations|journal = Proceedings of the IEEE|volume = 102|issue = 5|pages = 699–716|year = 2014|last1 = Benjamin|first1 = Ben Varkey|last2 = Peiran Gao|last3 = McQuinn|first3 = Emmett|last4 = Choudhary|first4 = Swadesh|last5 = Chandrasekaran|first5 = Anand R.|last6 = Bussat|first6 = Jean-Marie|last7 = Alvarez-Icaza|first7 = Rodrigo|last8 = Arthur|first8 = John V.|last9 = Merolla|first9 = Paul A.|last10 = Boahen|first10 = Kwabena|s2cid = 17176371}}</ref><br />
<br />
'''<font color="#ff8000">神经栅格Neurogrid<ref name=":15" /></font>'''是由斯坦福大学Brains in Silicon公司研发的、使用神经形态工程原理设计的硬件。该电路板由16个定制设计的芯片组成(NeuroCores)。在设计中，每个NeuroCore芯片的模拟电路对65536个神经元的神经元素进行模拟，以最大限度地提高能量效率。模拟出的神经元通过设计的数字电路连接，以最大化脉冲吞吐量。<ref name=":16" /><ref name=":17" /><br />
<br />
A research project with implications for neuromorphic engineering is the [[Human Brain Project]] that is attempting to simulate a complete human brain in a supercomputer using biological data. It is made up of a group of researchers in neuroscience, medicine, and computing.<ref name=":18">{{cite web|title=Involved Organizations|url=http://www.humanbrainproject.eu/partners.html|access-date=22 February 2013|url-status=dead|archive-url=https://web.archive.org/web/20130302142627/http://www.humanbrainproject.eu/partners.html|archive-date=2 March 2013}}</ref> [[Henry Markram]], the project's co-director, has stated that the project proposes to establish a foundation to explore and understand the brain and its diseases, and to use that knowledge to build new computing technologies. The three primary goals of the project are to better understand how the pieces of the brain fit and work together, to understand how to objectively diagnose and treat brain diseases, and to use the understanding of the human brain to develop neuromorphic computers. That the simulation of a complete human brain will require a supercomputer a thousand times more powerful than today's encourages the current focus on neuromorphic computers.<ref name=":19">{{cite web|title=Human Brain Project|url=http://www.humanbrainproject.eu|access-date=22 February 2013}}</ref> $1.3 billion has been allocated to the project by The [[European Commission]].<ref name=":20">{{cite web|title=The Human Brain Project and Recruiting More Cyberwarriors|url=http://www.marketplace.org/topics/tech/human-brain-project-and-recruiting-more-cyberwarriors|access-date=22 February 2013|date=January 29, 2013}}</ref><br />
<br />
'''<font color="#ff8000">人类大脑计划Human Brain Project</font>'''对神经形态工程具有较大影响，其主要任务是尝试用生物数据在超级计算机中模拟完整的人脑。人类大脑计划由神经科学、医学和计算机科学背景的研究人员组成。<ref name=":18" />该项目的联合主管亨利•马克拉姆(Henry Markram)表示，人类大脑计划的目的是建立一个探索和了解脑科学和脑疾病知识的基础，并利用这些知识来构建更先进的计算机技术。这个项目的三个主要目标分别是: 更好地理解大脑的各个部分是如何相互配合协同工作的; 理解如何客观地诊断和治疗脑部疾病; 以及利用对人类大脑的理解来开发神经形态计算机。模拟一个完整的人类大脑需要一台比现在强大一千倍的超级计算机，这不断激发着对神经形态计算机领域的研究兴趣。<ref name=":19" />欧盟委员会已经向人类大脑计划拨款13亿美元。<ref name=":20" /><br />
<br />
Other research with implications for neuromorphic engineering involves the [[BRAIN Initiative]]<ref name="economist">[https://www.economist.com/news/science-and-technology/21582495-computers-will-help-people-understand-brains-better-and-understanding-brains Neuromorphic computing: The machine of a new soul], The Economist, 2013-08-03</ref> and the [[TrueNorth]] chip from [[IBM]].<ref name=":21">{{cite journal|last1=Modha|first1=Dharmendra|title=A million spiking-neuron integrated circuit with a scalable communication network and interface|journal=Science|date=Aug 2014|volume=345|issue=6197|pages=668–673|doi=10.1126/science.1254642|pmid=25104385|bibcode=2014Sci...345..668M|s2cid=12706847}}</ref> Neuromorphic devices have also been demonstrated using nanocrystals, nanowires, and conducting polymers.<ref name=":22">{{Cite web|url=http://jessamynfairfield.com/wp-content/uploads/2017/03/PWMar17Fairfield.pdf|title=Smarter Machines|last=Fairfield|first=Jessamyn|date=March 1, 2017}}</ref><br />
<br />
其他与神经形态工程有关的研究还包括'''<font color="#ff8000">脑计划BRAIN initiative</font>'''，<ref name="economist" />和IBM研发的'''<font color="#ff8000">TrueNorth</font>'''芯片。<ref name=":21" />使用纳米晶体、纳米线和导电聚合物也能够用于制造神经形态学硬件。<ref name=":22" /><br />
<br />
[[Intel]] unveiled its neuromorphic research chip, called “[[Intel Loihi|Loihi]]”, in October 2017. The chip uses an asynchronous [[spiking neural network]] (SNN) to implement adaptive self-modifying event-driven fine-grained parallel computations used to implement learning and inference with high efficiency.<ref name=":23">{{cite journal |last1=Davies |first1=Mike |title=Loihi: A Neuromorphic Manycore Processor with On-Chip Learning |journal=IEEE Micro |date=January 16, 2018 |volume=38 |issue=1 |pages=82–99 |display-authors=etal|doi=10.1109/MM.2018.112130359 |s2cid=3608458 }}</ref><ref name="Morris2017">{{cite web |last1=Morris |first1=John |title=Why Intel built a neuromorphic chip |url=https://www.zdnet.com/article/why-intel-built-a-neuromorphic-chip/ |website=ZDNet |access-date=17 August 2018 |language=en}}</ref><br />
<br />
2017年10月，英特尔发布了神经形态芯片'''<font color="#ff8000">Loihi</font>'''。该芯片采用异步'''<font color="#ff8000">脉冲神经网络Spiking neural network</font>'''实现了自适应、自修改、事件驱动的细粒度并行计算，实现了高效的学习和推理。<ref name=":23" /><ref name="Morris2017" /><br />
<br />
[[IMEC]], a Belgium-based nanoelectronics research center, demonstrated the world's first self-learning neuromorphic chip. The brain-inspired chip, based on OxRAM technology, has the capability of self-learning and has been demonstrated to have the ability to compose music.<ref name=":24">{{cite web |title=Imec demonstrates self-learning neuromorphic chip that composes music |url=https://www.imec-int.com/en/articles/imec-demonstrates-self-learning-neuromorphic-chip-that-composes-music |website=IMEC International |access-date=1 October 2019}}</ref> IMEC released the 30-second tune composed by the prototype. The chip was sequentially loaded with songs in the same time signature and style. The songs were old Belgian and French flute minuets, from which the chip learned the rules at play and then applied them.<ref name=":25">{{cite web|last1=Bourzac|first1=Katherine|title=A Neuromorphic Chip That Makes Music|url=https://spectrum.ieee.org/a-neuromorphic-chip-that-makes-music|url-status=live|access-date=1 October 2019|website=IEEE Spectrum|date=May 23, 2017}}</ref><br />
<br />
比利时的'''<font color="#ff8000">微电子研究中心IMEC</font>'''研发了世界上首个自学习神经形态芯片。这种基于'''<font color="#32CD32"> OxRAM(filamentary-oxide-based resistive memory technology)</font>''' 技术的大脑启发芯片具有自学习能力，并且已被证明具有创作音乐的能力。<ref name=":24" />IMEC发布了由芯片原型机谱写的30秒旋律。向芯片加载一系列特征、风格统一的歌曲（古代比利时和法国长笛小步舞曲），芯片从中学习相关规则并将其应用于创作。<ref name=":25" /><br />
<br />
[[Blue Brain Project|The Blue Brain Project]], led by Henry Markram, aims to build biologically detailed digital reconstructions and simulations of the mouse brain. The Blue Brain Project has created in silico models of rodent brains, while attempting to replicate as many details about its biology as possible. The supercomputer-based simulations offer new perspectives on understanding the structure and functions of the brain.<br />
<br />
由亨利·马克拉姆领导的'''<font color="#ff8000">蓝脑计划The Blue Brain Project</font>'''旨在建立小鼠大脑生理学细节的数字重建和模拟。蓝脑计划已经建立了啮齿动物大脑的电子模型，同时进行着尽可能多地复制其生理学细节的尝试。基于超级计算机的模拟为理解大脑的结构和功能提供了新的视角。<br />
<br />
The European Union funded a series of projects at the University of Heidelberg, which led to the development of <br />
[[BrainScaleS]] (brain-inspired multiscale computation in neuromorphic hybrid systems), a hybrid analog [[neuromorphic]] supercomputer located at Heidelberg University, Germany. It was developed as part of the [[Human Brain Project]] neuromorphic computing platform and is the complement to the [[SpiNNaker]] supercomputer (which is based on digital technology). The architecture used in BrainScaleS mimics biological neurons and their connections on a physical level; additionally, since the components are made of silicon, these model neurons operate on average 864 times (24 hours of real time is 100 seconds in the machine simulation) that of their biological counterparts.<ref name=":26">{{Cite web|date=2016-03-21|title=Beyond von Neumann, Neuromorphic Computing Steadily Advances|url=https://www.hpcwire.com/2016/03/21/lacking-breakthrough-neuromorphic-computing-steadily-advance/|access-date=2021-10-08|website=HPCwire|language=en-US}}</ref><br />
<br />
欧盟资助了海德堡大学一系列促进BrainScaleS(神经形态混合系统中受大脑启发的多尺度计算)发展的项目，这是一台位于德国海德堡大学的混合模拟'''<font color="#ff8000">神经形态Neuromorphic</font>'''超级计算机。它是作为人类大脑计划中神经形态计算平台的一部分而开发的，是'''<font color="#ff8000">SpiNNaker</font>'''超级计算机(基于数字技术)的补充。BrainScaleS中使用的体系架构模拟了生物神经元及其在物理层面上的连接；此外，由于这些组件是由硅制成的，这些模型神经元平均运行速度是生物神经元的864倍，这意味着在机器模拟中，24小时的实时时间仅为100秒。<ref name=":26" /><br />
<br />
===Neuromorphic sensors===<br />
===神经形态传感器===<br />
The concept of neuromorphic systems can be extended to sensors (not just to computation). An example of this applied to detecting [[light]] is the [[retinomorphic sensor]] or, when employed in an array, the [[event camera]].<br />
<br />
神经形态系统的概念可以扩展到传感器(而不仅仅是计算单元)。一个用于检测光线的例子是'''<font color="#ff8000">类视网膜传感器Retinomorphic sensor</font>'''，或者'''<font color="#ff8000">事件摄像机Event camera</font>阵列。'''<br />
<br />
<br />
==Ethical considerations ==<br />
==伦理问题==<br />
While the interdisciplinary concept of neuromorphic engineering is relatively new, many of the same ethical considerations apply to neuromorphic systems as apply to [[human-like machines]] and [[artificial intelligence]] in general. However, the fact that neuromorphic systems are designed to mimic a [[human brain]] gives rise to unique ethical questions surrounding their usage.<br />
<br />
虽然神经形态工程这个跨学科概念相对较新，但许多适用于'''<font color="#ff8000">类人机器Human-like machines</font>'''和人工智能的伦理考虑大体上也适用于神经形态系统。另一方面，神经形态系统是为了模仿人类大脑而设计，这一事实引起了一些针对性的独特伦理问题。<br />
<br />
However, the practical debate is that neuromorphic hardware as well as artificial "neural networks" are immensely simplified models of how the brain operates or processes information at a much lower [[complex system|complexity in terms of size and functional technology]] and a much more regular structure in terms of [[brain connectivity|connectivity]]. Comparing [[neuromorphic chip]]s to the brain is a very crude comparison similar to comparing a plane to a bird just because they both have wings and a tail. The fact is that neural cognitive systems are many orders of magnitude more [[Energy efficiency (physics)|energy-]] and compute-efficient than current state-of-the-art AI and neuromorphic engineering is an attempt to narrow this gap by inspiring from the brain's mechanism just like many engineering designs have [[bioengineering|bio-inspired features]].<br />
<br />
然而，这种争论的实际情况是，神经形态硬件和人工“神经网络”是对大脑运作或处理信息过程的极其简化的模型，在大小和功能技术方面的'''<font color="#ff8000">复杂性Complexity</font>'''相比而言比较低，在'''<font color="#ff8000">连接性Connectivity</font>'''方面也具有更加规则的结构。将'''<font color="#ff8000">神经形态芯片Neuromorphic chips</font>'''与大脑进行比较是一种非常粗糙的比较，类似于仅仅因为飞机有翅膀和尾巴就将它与鸟进行比较。事实上，当前最先进的人工智能在'''<font color="#ff8000">能耗效率Energy-efficiency</font>'''和计算效率方面距离人脑神经认知系统仍有较大差距，而神经形态工程只是一种通过从大脑机制中获得灵感来缩小这种差距的尝试，就像许多工程设计中都具有'''<font color="#ff8000">生物启发的特征Bio-inspired features</font>'''一样。<br />
<br />
===Democratic concerns===<br />
===公众担忧===<br />
Significant ethical limitations may be placed on neuromorphic engineering due to public perception.<ref name=":27">{{Cite report|url=https://ai100.stanford.edu/sites/g/files/sbiybj9861/f/ai_100_report_0831fnl.pdf|title=Artificial Intelligence and Life in 2030|author=2015 Study Panel|date=September 2016|work=One Hundred Year Study on Artificial Intelligence (AI100)|publisher=Stanford University}}</ref> Special [[Eurobarometer]] 382: Public Attitudes Towards Robots, a survey conducted by the European Commission, found that 60% of [[European Union]] citizens wanted a ban of robots in the care of children, the elderly, or the disabled. Furthermore, 34% were in favor of a ban on robots in education, 27% in healthcare, and 20% in leisure. The European Commission classifies these areas as notably “human.” The report cites increased public concern with robots that are able to mimic or replicate human functions. Neuromorphic engineering, by definition, is designed to replicate the function of the human brain.<ref name=":1">{{Cite web|url=http://ec.europa.eu/commfrontoffice/publicopinion/archives/ebs/ebs_382_en.pdf|title=Special Eurobarometer 382: Public Attitudes Towards Robots|last=European Commission|date=September 2012|website=European Commission}}</ref><br />
<br />
由于公众认知的相关忧虑，神经形态工程学可能会受到严重的伦理限制。<ref name=":27" />欧盟委员会进行的一项调查发现，60% 的欧盟公民希望禁止机器人参与照顾儿童、老人或残疾人的工作。此外，34% 的人支持禁止机器人用于教育，27% 的人支持禁止机器人用于医疗保健，20% 的人支持禁止机器人用于休闲。欧盟委员会将以上领域划入“人类”范畴。报告指出，公众越来越关注能够模仿或复制人类功能的机器人。神经形态工程，顾名思义，是为了复制人脑的功能而设计的。<ref name=":1" /><br />
<br />
The democratic concerns surrounding neuromorphic engineering are likely to become even more profound in the future. The European Commission found that EU citizens between the ages of 15 and 24 are more likely to think of robots as human-like (as opposed to instrument-like) than EU citizens over the age of 55. When presented an image of a robot that had been defined as human-like, 75% of EU citizens aged 15–24 said it corresponded with the idea they had of robots while only 57% of EU citizens over the age of 55 responded the same way. The human-like nature of neuromorphic systems, therefore, could place them in the categories of robots many EU citizens would like to see banned in the future.<ref name=":1" /><br />
<br />
围绕神经形态工程的公众担忧可能在未来变得更加深刻。欧盟委员会发现，相比于55岁以上的欧盟公民，15至24岁的欧盟公民更有可能认为机器人像人(而不是像仪器)。当看到一张“类人”机器人的图片时，年龄在15岁至24岁之间的欧盟公民中有75% 的人表示这符合他们对机器人的想法，而55岁以上的欧盟公民中只有57% 的人有同样的反应。因此，神经形态系统可能因为其类似人类的特性而被归入许多欧盟公民希望在未来禁止使用的机器人类别。<ref name=":1" /><br />
<br />
=== Personhood===<br />
=== 人格问题===<br />
As neuromorphic systems have become increasingly advanced, some scholars{{who|date=August 2021}} have advocated for granting [[personhood]] rights to these systems. If the brain is what grants humans their personhood, to what extent does a neuromorphic system have to mimic the human brain to be granted personhood rights? Critics of technology development in the [[Human Brain Project]], which aims to advance brain-inspired computing, have argued that advancement in neuromorphic computing could lead to machine consciousness or personhood.<ref name=":28">{{Cite journal|last=Aicardi|first=Christine|date=September 2018|title=Accompanying technology development in the Human Brain Project: From foresight to ethics management|journal=Futures|volume=102|pages=114–124|doi=10.1016/j.futures.2018.01.005|doi-access=free}}</ref> If these systems are to be treated as people, critics argue, then many tasks humans perform using neuromorphic systems, including the act of termination of neuromorphic systems, may be morally impermissible as these acts would violate the autonomy of the neuromorphic systems.<ref name=":29">{{Cite journal|last=Lim|first=Daniel|date=2014-06-01|title=Brain simulation and personhood: a concern with the Human Brain Project|journal=Ethics and Information Technology|language=en|volume=16|issue=2|pages=77–89|doi=10.1007/s10676-013-9330-5|s2cid=17415814|issn=1572-8439}}</ref><br />
<br />
随着神经形态系统的日益发展，一些学者主张赋予这些系统'''<font color="ff8000">人格Personhood</font>'''权。如果是大脑赋予了人类人格，那么在多大程度上模仿人类大脑的神经形态系统才能被赋予人格权利？“人类大脑计划”旨在推进以大脑为灵感的计算机技术发展，该计划的批评者认为，神经形态计算机技术的进步可能导致机器意识或人格的形成。<ref name=":28" />这些批评者认为，如果这些系统被当作人来对待，那么人类使用神经形态系统执行任务（包括终止神经形态系统）的行为，在道德上就可能是不被允许的，因为这些行为将违反神经形态系统的自主性。<ref name=":29" /><br />
<br />
<br />
==Dual use (military applications)==<br />
==军民两用技术（军事应用）==<br />
The [[Joint Artificial Intelligence Center]], a branch of the U.S. military, is a center dedicated to the procurement and implementation of AI software and neuromorphic hardware for combat use. Specific applications include smart headsets/goggles and robots. JAIC intends to rely heavily on neuromorphic technology to connect "every fighter every shooter" within a network of neuromorphic-enabled units.<br />
<br />
'''<font color="#ff8000">联合人工智能中心The Joint Artificial Intelligence Center</font>'''(JAIC),是美国军队的一个分支，专门从事采购和实施用于战斗的人工智能软件和神经形态硬件。具体应用包括智能耳机/护目镜和机器人。JAIC打算高度依赖神经形态技术，'''<font color="#32CD32">使用神经形态技术来连接神经形态单位网络中的“每个战士每个射手”</font>'''。<br />
<br />
<br />
==Legal considerations==<br />
==法律问题==<br />
Skeptics have argued that there is no way to apply the electronic personhood, the concept of personhood that would apply to neuromorphic technology, legally. In a letter signed by 285 experts in law, robotics, medicine, and ethics opposing a European Commission proposal to recognize “smart robots” as legal persons, the authors write, “A legal status for a robot can’t derive from the [[Natural person|Natural Person]] model, since the robot would then hold [[human rights]], such as the right to dignity, the right to its integrity, the right to remuneration or the right to citizenship, thus directly confronting the Human rights. This would be in contradiction with the [[Charter of Fundamental Rights of the European Union]] and the [[Convention for the Protection of Human Rights and Fundamental Freedoms]].”<ref>{{Cite web|url=http://www.robotics-openletter.eu/|title=Robotics Openletter {{!}} Open letter to the European Commission|language=fr-FR|access-date=2019-05-10}}</ref><br />
<br />
Skeptics have argued that there is no way to apply the electronic personhood, the concept of personhood that would apply to neuromorphic technology, legally. In a letter signed by 285 experts in law, robotics, medicine, and ethics opposing a European Commission proposal to recognize “smart robots” as legal persons, the authors write, “A legal status for a robot can’t derive from the Natural Person model, since the robot would then hold human rights, such as the right to dignity, the right to its integrity, the right to remuneration or the right to citizenship, thus directly confronting the Human rights. This would be in contradiction with the Charter of Fundamental Rights of the European Union and the Convention for the Protection of Human Rights and Fundamental Freedoms.”<br />
<br />
怀疑论者认为，在法律上没有办法应用电子人格，这个人格概念将适用于神经形态技术。在一封由285名法律、机器人、医学和伦理学专家签名的信中，作者们反对欧盟委员会承认“智能机器人”为法人的提议，他们写道，“机器人的法律地位不能从自然人模型中推导出来，因为机器人将拥有人权，如尊严权、完整权、报酬权或公民权，从而直接面对人权。这将有悖于《欧洲联盟基本权利宪章和《保护人权和基本自由公约》。”<br />
<br />
===Ownership and property rights===<br />
===所有权及财产权问题===<br />
There is significant legal debate around property rights and artificial intelligence. In ''Acohs Pty Ltd v. Ucorp Pty Ltd'', Justice Christopher Jessup of the [[Federal Court of Australia]] found that the [[source code]] for [[Material safety data sheets|Material Safety Data Sheets]] could not be [[Copyright law of Australia|copyrighted]] as it was generated by a [[software interface]] rather than a human author.<ref>{{Cite web|url=http://www.lavan.com.au/advice/intellectual_property_technology/copyright_in_source_code_and_digital_products|title=Copyright in source code and digital products|last=Lavan|website=Lavan|language=en|access-date=2019-05-10}}</ref> The same question may apply to neuromorphic systems: if a neuromorphic system successfully mimics a human brain and produces a piece of original work, who, if anyone, should be able to claim ownership of the work?<ref>{{cite journal |last1=Eshraghian|first1=Jason K. |title=Human Ownership of Artificial Creativity |journal=Nature Machine Intelligence |date=9 March 2020 |volume=2 |pages=157–160 |doi=10.1038/s42256-020-0161-x}}</ref><br />
<br />
There is significant legal debate around property rights and artificial intelligence. In Acohs Pty Ltd v. Ucorp Pty Ltd, Justice Christopher Jessup of the Federal Court of Australia found that the source code for Material Safety Data Sheets could not be copyrighted as it was generated by a software interface rather than a human author. The same question may apply to neuromorphic systems: if a neuromorphic system successfully mimics a human brain and produces a piece of original work, who, if anyone, should be able to claim ownership of the work?<br />
<br />
围绕财产权和人工智能有着重大的法律争论。在 Acohs Pty Ltd 诉 Ucorp Pty Ltd 一案中，澳大利亚联邦法院的克里斯托弗 · 杰瑟普法官发现，材料安全数据表的源代码不能受版权保护，因为它是由软件界面而不是人工作者生成的。同样的问题可能也适用于神经形态系统: 如果一个神经形态系统成功地模仿了人类的大脑并产生了一部原创作品，那么谁，如果有人，应该声称拥有这部作品的所有权？<br />
<br />
==Neuromemristive systems==<br />
==神经记忆电阻系统 ==<br />
Neuromemristive systems are a subclass of neuromorphic computing systems that focus on the use of [[memristors]] to implement [[neuroplasticity]]. While neuromorphic engineering focuses on mimicking biological behavior, neuromemristive systems focus on abstraction.<ref>{{Cite web|url=https://digitalops.sandia.gov/Mediasite/Play/a10cf6ceb55d47608bb8326dd00e46611d|title=002.08 N.I.C.E. Workshop 2014: Towards Intelligent Computing with Neuromemristive Circuits and Systems - Feb. 2014|website=digitalops.sandia.gov|access-date=2019-08-26}}</ref> For example, a neuromemristive system may replace the details of a [[Cerebral cortex|cortical]] microcircuit's behavior with an abstract neural network model.<ref>C. Merkel and D. Kudithipudi, "Neuromemristive extreme learning machines for pattern classification," ISVLSI, 2014.</ref><br />
<br />
Neuromemristive systems are a subclass of neuromorphic computing systems that focus on the use of memristors to implement neuroplasticity. While neuromorphic engineering focuses on mimicking biological behavior, neuromemristive systems focus on abstraction. For example, a neuromemristive system may replace the details of a cortical microcircuit's behavior with an abstract neural network model.C. Merkel and D. Kudithipudi, "Neuromemristive extreme learning machines for pattern classification," ISVLSI, 2014.<br />
<br />
神经记忆电阻系统是神经形态计算系统的一个亚类，主要研究利用记忆电阻器实现神经可塑性。神经形态工程的重点是模拟生物行为，而神经记忆电阻系统的重点是提取。例如，一个神经记忆系统可以用一个抽象的神经网络模型取代皮层微电路的行为细节。默克尔和 d. Kudithipudi，“用于模式分类的神经记忆极端学习机器”，ISVLSI，2014。<br />
<br />
There exist several neuron inspired threshold logic functions<ref name="Maan 1–13" /> implemented with memristors that have applications in high level [[pattern recognition]] applications. Some of the applications reported recently include [[speech recognition]],<ref>{{Cite journal|title = Memristor pattern recogniser: isolated speech word recognition|journal = Electronics Letters|pages = 1370–1372|volume = 51|issue = 17|doi = 10.1049/el.2015.1428|first1 = A.K.|last1 = Maan|first2 = A.P.|last2 = James|first3 = S.|last3 = Dimitrijev|year = 2015|bibcode = 2015ElL....51.1370M|hdl = 10072/140989|s2cid = 61454815|url = https://semanticscholar.org/paper/48d3ab11ec6e213b62f11eedcfb7b7febb058674|hdl-access = free}}</ref> [[face recognition]]<ref>{{Cite journal|title = Memristive Threshold Logic Face Recognition|journal = Procedia Computer Science|date = 2014-01-01|pages = 98–103|volume = 41|series = 5th Annual International Conference on Biologically Inspired Cognitive Architectures, 2014 BICA|doi = 10.1016/j.procs.2014.11.090|first1 = Akshay Kumar|last1 = Maan|first2 = Dinesh S.|last2 = Kumar|first3 = Alex Pappachen|last3 = James|doi-access = free}}</ref> and [[object recognition]].<ref>{{Cite journal|title = Memristive Threshold Logic Circuit Design of Fast Moving Object Detection|journal = IEEE Transactions on Very Large Scale Integration (VLSI) Systems|date = 2015-10-01|issn = 1063-8210|pages = 2337–2341|volume = 23|issue = 10|doi = 10.1109/TVLSI.2014.2359801|first1 = A.K.|last1 = Maan|first2 = D.S.|last2 = Kumar|first3 = S.|last3 = Sugathan|first4 = A.P.|last4 = James|arxiv = 1410.1267|s2cid = 9647290}}</ref> They also find applications in replacing conventional digital logic gates.<ref>{{Cite journal|title = Resistive Threshold Logic|journal = IEEE Transactions on Very Large Scale Integration (VLSI) Systems|date = 2014-01-01|issn = 1063-8210|pages = 190–195|volume = 22|issue = 1|doi = 10.1109/TVLSI.2012.2232946|first1 = A.P.|last1 = James|first2 = L.R.V.J.|last2 = Francis|first3 = D.S.|last3 = Kumar|arxiv = 1308.0090|s2cid = 7357110}}</ref><ref>{{Cite journal|title = Threshold Logic Computing: Memristive-CMOS Circuits for Fast Fourier Transform and Vedic Multiplication|journal = IEEE Transactions on Very Large Scale Integration (VLSI) Systems|date = 2015-11-01|issn = 1063-8210|pages = 2690–2694|volume = 23|issue = 11|doi = 10.1109/TVLSI.2014.2371857|first1 = A.P.|last1 = James|first2 = D.S.|last2 = Kumar|first3 = A.|last3 = Ajayan|arxiv = 1411.5255|s2cid = 6076956}}</ref><br />
<br />
There exist several neuron inspired threshold logic functions implemented with memristors that have applications in high level pattern recognition applications. Some of the applications reported recently include speech recognition, face recognition and object recognition. They also find applications in replacing conventional digital logic gates.<br />
<br />
利用记忆电阻器实现的神经元启发阈值逻辑函数在高级模式识别中有着广泛的应用。最近报道的一些应用包括语音识别、人脸识别和物体识别。它们还可以用来取代传统的数字逻辑门。<br />
<br />
For ideal passive memristive circuits there is an exact equation (Caravelli-Traversa-[[Di Ventra]] equation) for the internal memory of the circuit:<ref>{{cite journal |last=Caravelli |display-authors=etal|arxiv=1608.08651 |title=The complex dynamics of memristive circuits: analytical results and universal slow relaxation |year=2017 |doi=10.1103/PhysRevE.95.022140 |pmid= 28297937 |volume=95 |issue= 2 |pages= 022140 |journal=Physical Review E|bibcode=2017PhRvE..95b2140C |s2cid=6758362}}</ref><br />
<br />
For ideal passive memristive circuits there is an exact equation (Caravelli-Traversa-Di Ventra equation) for the internal memory of the circuit:<br />
<br />
对于理想的无源记忆电路，电路的内部记忆有一个精确的方程(Caravelli-Traversa-Di Ventra 方程) :<br />
<br />
:<math> \frac{d}{dt} \vec{W} = \alpha \vec{W}-\frac{1}{\beta} (I+\xi \Omega W)^{-1} \Omega \vec S </math><br />
<br />
:\frac{d}{dt} \vec{W} = \alpha \vec{W}-\frac{1}{\beta} (I+\xi \Omega W)^{-1} \Omega \vec S<br />
<br />
:\frac{d}{dt} \vec{W} = \alpha \vec{W}-\frac{1}{\beta} (I+\xi \Omega W)^{-1} \Omega \vec S<br />
<br />
as a function of the properties of the physical memristive network and the external sources. In the equation above, <math>\alpha</math> is the "forgetting" time scale constant, <math> \xi=r-1</math> and <math>r=\frac{R_\text{off}}{R_\text{on}}</math> is the ratio of ''off'' and ''on'' values of the limit resistances of the memristors, <math> \vec S </math> is the vector of the sources of the circuit and <math>\Omega</math> is a projector on the fundamental loops of the circuit. The constant <math>\beta</math> has the dimension of a voltage and is associated to the properties of the [[memristor]]; its physical origin is the charge mobility in the conductor. The diagonal matrix and vector <math>W=\operatorname{diag}(\vec W)</math> and <math>\vec W</math> respectively, are instead the internal value of the memristors, with values between 0 and 1. This equation thus requires adding extra constraints on the memory values in order to be reliable.<br />
<br />
<nowiki>as a function of the properties of the physical memristive network and the external sources. In the equation above, \alpha is the "forgetting" time scale constant, \xi=r-1 and r=\frac{R_\text{off}}{R_\text{on}} is the ratio of off and on values of the limit resistances of the memristors, \vec S is the vector of the sources of the circuit and \Omega is a projector on the fundamental loops of the circuit. The constant \beta has the dimension of a voltage and is associated to the properties of the memristor; its physical origin is the charge mobility in the conductor. The diagonal matrix and vector W=\operatorname{diag}(\vec W) and \vec W respectively, are instead the internal value of the memristors, with values between 0 and 1. This equation thus requires adding extra constraints on the memory values in order to be reliable.</nowiki><br />
<br />
<nowiki>作为物理记忆网络和外部源的性质的函数。在上述方程中，α 是“遗忘”时间尺度常数，xi = r-1，r = frac { r _ text { off }{ on }{ r _ text { on }}是记忆电阻器极限电阻的开关和开关值之比，vec s 是电路源的矢量，Omega 是电路基本环路的投影仪。常数 β 具有电压的尺寸，与记忆电阻器的特性有关; 它的物理起源是导体中的电荷迁移率。对角矩阵和向量 w = 操作者名{ diag }(vec w)和 vec w 分别是记忆电阻器的内值，值在0到1之间。因此，这个等式需要在内存值上添加额外的约束，以保证可靠性。</nowiki><br />
<br />
==See also==<br />
==相关词条==<br />
{{Columns-list|colwidth=18em|<br />
* [[AI accelerator (computer hardware)|AI accelerator]]<br />
* [[Artificial brain]]<br />
* [[Biomorphic]]<br />
* [[Cognitive computer]]<br />
* [[Computation and Neural Systems]]<br />
* [[Differentiable programming]]<br />
* [[Event camera]]<br />
* [[Neurorobotics]]<br />
* [[Optical flow sensor]]<br />
* [[Physical neural network]]<br />
* [[SpiNNaker]]<br />
* [[SyNAPSE]]<br />
* [[Retinomorphic sensor]]<br />
* [[Vision chip]]<br />
* [[Vision processing unit]]<br />
* [[Zeroth (software)]]<br />
* [[Hardware for artificial intelligence]]<br />
}}<br />
<br />
{{Portal bar|Electronics}}<br />
<br />
== References==<br />
{{Reflist|40em}}<br />
<br />
==External links==<br />
<!--======================== {{No more links}} ============================<br />
| PLEASE BE CAUTIOUS IN ADDING MORE LINKS TO THIS ARTICLE. Wikipedia |<br />
| is not a collection of links nor should it be used for advertising. |<br />
| |<br />
| Excessive or inappropriate links WILL BE DELETED. |<br />
| See [[Wikipedia:External links]] & [[Wikipedia:Spam]] for details. |<br />
| |<br />
| If there are already plentiful links, please propose additions or |<br />
| replacements on this article's discussion page, or submit your link |<br />
| to the relevant category at the Open Directory Project (dmoz.org) |<br />
| and link back to that category using the {{dmoz}} template. |<br />
======================={{No more links}}=============================--><br />
{{Commons category|Neuromorphic Engineering}}<br />
*[https://web.archive.org/web/20150727034331/http://ine-web.org/workshops/workshops-overview Telluride Neuromorphic Engineering Workshop]<br />
*[https://archive.today/20130115190057/http://capocaccia.ethz.ch/ CapoCaccia Cognitive Neuromorphic Engineering Workshop]<br />
*[https://web.archive.org/web/20190716132350/http://www.ine-web.org/ Institute of Neuromorphic Engineering]<br />
*[http://www.ine-news.org/ INE news site].<br />
*[http://www.frontiersin.org/neuromorphic_engineering Frontiers in Neuromorphic Engineering Journal]<br />
*[http://www.cns.caltech.edu/ Computation and Neural Systems] department at the [[California Institute of Technology]].<br />
*[http://www.humanbrainproject.eu/ Human Brain Project official site]<br />
*[https://www.the-scientist.com/features/building-a-silicon-brain-65738 Building a Silicon Brain:] Computer chips based on biological neurons may help simulate larger and more-complex brain models. May 1, 2019. SANDEEP RAVINDRAN<br />
<br />
<br />
*Telluride Neuromorphic Engineering Workshop<br />
*CapoCaccia Cognitive Neuromorphic Engineering Workshop<br />
*Institute of Neuromorphic Engineering<br />
*INE news site.<br />
*Frontiers in Neuromorphic Engineering Journal<br />
*Computation and Neural Systems department at the California Institute of Technology.<br />
*Human Brain Project official site<br />
*Building a Silicon Brain: Computer chips based on biological neurons may help simulate larger and more-complex brain models. May 1, 2019. SANDEEP RAVINDRAN<br />
<br />
=<nowiki>外部链接</nowiki>= <br />
*碲化物神经形态工程工作室<br />
*CapoCaccia 认知神经形态工程工作室<br />
*神经形态工程研究所<br />
*INE 新闻站点。<br />
*《神经形态工程学前沿》<br />
*加州理工学院计算与神经系统系。<br />
*人脑项目官方网站<br />
*打造硅脑: 基于生物神经元的计算机芯片可能有助于模拟更大、更复杂的大脑模型。2019年5月1日。SANDEEP RAVINDRAN<br />
<br />
{{Differentiable computing}}<br />
{{Authority control}}<br />
<br />
[[Category:Electrical engineering]]<br />
[[Category:Neuroscience]]<br />
[[分类:AI accelerators|*]]<br />
[[Category:Artificial intelligence]]<br />
[[Category:Robotics]]<br />
<br />
Category:Electrical engineering<br />
Category:Neuroscience<br />
*<br />
Category:Artificial intelligence<br />
Category:Robotics<br />
<br />
类别: 电气工程类别: 神经科学 * 类别: 人工智能类别: 机器人<br />
<br />
<noinclude><br />
<br />
<small>This page was moved from [[wikipedia:en:Neuromorphic engineering]]. Its edit history can be viewed at [[神经形态计算/edithistory]]</small></noinclude><br />
<br />
[[Category:待整理页面]]<br />
<br />
<references /></div>Shenky20https://wiki.swarma.org/index.php?title=%E7%A5%9E%E7%BB%8F%E5%BD%A2%E6%80%81%E8%AE%A1%E7%AE%97&diff=30746神经形态计算2022-04-28T10:35:11Z<p>Shenky20：/* 伦理问题 */</p>
<hr />
<div><br />
此词条由神经动力学模型读书会词条梳理志愿者Shenky20翻译审校，未经专家审校，带来阅读不便，请见谅。<br />
<br />
{{Use American English|date = January 2019}}<br />
{{Use mdy dates|date = January 2019}}<br />
<br />
'''Neuromorphic engineering''', also known as '''neuromorphic computing''',<ref name=":3">{{Cite journal | doi = 10.1145/2601069| title = Neuromorphic computing gets ready for the (really) big time| journal = [[Communications of the ACM]]| volume = 57| issue = 6| year = 2014| last1 = Monroe | first1 = D. | pages = 13–15| s2cid = 20051102}}</ref><ref name=":4">{{Cite journal | doi = 10.1088/0957-4484/21/17/175202| pmid = 20368686| title = Nanotube devices based crossbar architecture: Toward neuromorphic computing| journal = Nanotechnology| volume = 21| issue = 17| pages = 175202| year = 2010| last1 = Zhao | first1 = W. S. | last2 = Agnus | first2 = G. | last3 = Derycke | first3 = V. | last4 = Filoramo | first4 = A. | last5 = Bourgoin | first5 = J. -P. | last6 = Gamrat | first6 = C. | bibcode = 2010Nanot..21q5202Z| url = https://zenodo.org/record/3428659}}</ref><ref name="humanbrainproject">{{YouTube|id=6RoiZ90mGfw|title=The Human Brain Project SP 9: Neuromorphic Computing Platform}}</ref> is the use of [[very-large-scale integration]] (VLSI) systems containing electronic [[analog circuit]]s to mimic neuro-biological architectures present in the nervous system. A neuromorphic computer/chip is any device that uses physical artificial neurons (made from silicon) to do computations.<ref name=":5">{{cite journal|last1=Mead|first1=Carver|title=Neuromorphic electronic systems|journal=Proceedings of the IEEE|date=1990|volume=78|issue=10|pages=1629–1636|doi=10.1109/5.58356|url=https://authors.library.caltech.edu/53090/1/00058356.pdf}}</ref><ref name=":2" /> In recent times, the term ''neuromorphic'' has been used to describe analog, digital, mixed-mode analog/digital VLSI, and software systems that implement models of [[neural system]]s (for [[perception]], [[motor control]], or [[multisensory integration]]). The implementation of neuromorphic computing on the hardware level can be realized by oxide-based [[memristor]]s,<ref name="Maan 1–13">{{Cite journal|last1=Maan|first1=A. K.|last2=Jayadevi|first2=D. A.|last3=James|first3=A. P.|date=2016-01-01|title=A Survey of Memristive Threshold Logic Circuits|journal=IEEE Transactions on Neural Networks and Learning Systems|volume=PP|issue=99|pages=1734–1746|doi=10.1109/TNNLS.2016.2547842|pmid=27164608|issn=2162-237X|arxiv=1604.07121|bibcode=2016arXiv160407121M|s2cid=1798273}}</ref> spintronic memories, threshold switches, and [[transistor]]s.<ref name=":6">{{Cite journal|title = Mott Memory and Neuromorphic Devices|journal = Proceedings of the IEEE|date = 2015-08-01|issn = 0018-9219|pages = 1289–1310|volume = 103|issue = 8|doi = 10.1109/JPROC.2015.2431914|first1 = You|last1 = Zhou|first2 = S.|last2 = Ramanathan|s2cid = 11347598|url = https://zenodo.org/record/895565}}</ref><ref name=":2">{{Cite document|title=Neuromorphic Circuits With Neural Modulation Enhancing the Information Content of Neural Signaling {{!}} International Conference on Neuromorphic Systems 2020|language=EN|doi=10.1145/3407197.3407204|s2cid=220794387}}</ref> Training software-based neuromorphic systems of spiking neural networks can be achieved using error backpropagation, e.g., using [[Python (programming language)|Python]] based frameworks such as snnTorch,<ref name=":7">{{cite journal |last1=Eshraghian|first1=Jason K.|last2=Ward|first2=Max|last3=Neftci |first3=Emre|last4=Wang|first4=Xinxin|last5=Lenz|first5=Gregor|last6=Dwivedi|first6=Girish|last7=Bennamoun|first7=Mohammed|last8=Jeong|first8=Doo Seok|last9=Lu|first9=Wei D.|title=Training Spiking Neural Networks Using Lessons from Deep Learning |date=1 October 2021 |arxiv=2109.12894 }}</ref> or using canonical learning rules from the biological learning literature, e.g., using BindsNet.<ref name=":8">{{Cite web | url=https://github.com/Hananel-Hazan/bindsnet | title=Hananel-Hazan/bindsnet: Simulation of spiking neural networks (SNNs) using PyTorch.| date=31 March 2020}}</ref><br />
<br />
<br />
'''<font color="#ff8000">神经形态工程Neuromorphic engineering</font>'''(也称为'''<font color="#ff8000">神经形态计算Neuromorphic computing</font>''')<ref name=":3" /><ref name=":4" /><ref name="humanbrainproject" />是指使用包含电子'''<font color="#ff8000">模拟电路Analog circuit</font>'''的'''<font color="#ff8000">超大规模集成电路Very-large-scale integration</font>'''系统来模拟神经系统中生理结构的研究方法。神经形态计算机或神经形态芯片包括任何使用由硅制成的人造神经元进行计算的设备。<ref name=":5" /><ref name=":2" />近年来，神经形态学（neuromorphic）这个术语被用来描述能够实现'''<font color="#ff8000">神经系统Neural system</font>'''模型功能(如'''<font color="#ff8000">感知Perception</font>'''、'''<font color="#ff8000">运动控制Motor control</font>'''，'''<font color="#ff8000">多感官整合Multisensory integration</font>'''等)的模拟、数字、模拟/数字混合模式超大规模集成电路和软件系统。神经形态计算的硬件实现可以通过基于氧化物的'''<font color="#ff8000">记忆电阻器Memristor</font>'''、<ref name="Maan 1–13" />自旋电子存储器、阈值开关和'''<font color="#ff8000">晶体管Transistor</font>'''来实现。<ref name=":6" /><ref name=":2" />对基于软件的脉冲神经网络系统的训练可以通过误差反向传播机制来实现，例如，使用snnTorch等基于Python的框架，<ref name=":7" />或使用BindsNet等典型的受生物启发的学习模式。<ref name=":8" /><br />
<br />
A key aspect of neuromorphic engineering is understanding how the morphology of individual neurons, circuits, applications, and overall architectures creates desirable computations, affects how information is represented, influences robustness to damage, incorporates learning and development, adapts to local change (plasticity), and facilitates evolutionary change.<br />
<br />
<br />
神经形态工程领域的一个关键问题，就是理解'''<font color="#32CD32">单个神经元形态、神经回路、应用和整体结构</font>'''如何产生理想的计算，如何影响信息的表示和对破坏的鲁棒性，如何整合学习和发展，如何适应局部变化(可塑性) 并促进逐渐发展的变化。<br />
<br />
Neuromorphic engineering is an interdisciplinary subject that takes inspiration from [[biology]], [[physics]], [[mathematics]], [[computer science]], and [[electronic engineering]]<ref name=":2" /> to design artificial neural systems, such as [[Machine vision|vision systems]], head-eye systems, auditory processors, and autonomous robots, whose physical architecture and design principles are based on those of biological nervous systems.<ref name=":9">{{Cite journal | doi = 10.1155/2012/705483| title = Qualitative Functional Decomposition Analysis of Evolved Neuromorphic Flight Controllers| journal = Applied Computational Intelligence and Soft Computing| volume = 2012| pages = 1–21| year = 2012| last1 = Boddhu | first1 = S. K. | last2 = Gallagher | first2 = J. C. | doi-access = free}}</ref> It was developed by [[Carver Mead]]<ref name=":10">{{cite web|title=carver mead website|last1=Mead|first1=Carver|url=http://www.carvermead.caltech.edu/index.html|website=carvermead}}</ref> in the late 1980s.<br />
<br />
<br />
神经形态工程是以'''<font color="#ff8000">生物学Biology</font>'''、'''<font color="#ff8000">物理学Physics</font>'''、'''<font color="#ff8000">数学Mathematics</font>'''、'''<font color="#ff8000">计算机科学Computer science</font>'''和'''<font color="#ff8000">电子工程Electronic engineering<ref name=":2" /></font>'''等学科为基础，设计人工神经系统（如'''<font color="#ff8000">视觉系统Vision systems</font>'''、头眼系统、听觉处理器以及物理结构和设计原则都受启发于生物神经系统的自主机器人）的一门交叉学科。<ref name=":9" />20世纪80年代后期，卡弗·米德极大地推动了神经形态工程领域的发展。<ref name=":10" /><br />
<br />
==Neurological inspiration==<br />
==神经生物学启发==<br />
Neuromorphic engineering is set apart by the inspiration it takes from what we know about the structure and operations of the [[brain]]. Neuromorphic engineering translates what we know about the brain's function into computer systems. Work has mostly focused on replicating the analog nature of [[biological computation]] and the role of [[neuron]]s in [[cognition]].<br />
<br />
<br />
神经形态工程形成的灵感来源于现已知的'''<font color="#ff8000">大脑Brain</font>'''结构及运作机制知识，它将我们对大脑功能的了解用于研究和优化计算机系统。该领域工作主要集中于对'''<font color="#ff8000">生物计算Biological computation</font>'''模拟特性和'''<font color="#ff8000">神经元Neuron</font>'''在'''<font color="#ff8000">认知Cognition</font>'''中所发挥作用的复刻。<br />
<br />
The biological processes of neurons and their [[synapse]]s are dauntingly complex, and thus very difficult to artificially simulate. A key feature of biological brains is that all of the processing in neurons use analog chemical signals. This makes it hard to replicate brains in computers because the current generation of computers is completely digital. However, the characteristics of these parts can be abstracted into mathematical functions that closely capture the essence of the neuron's operations.<br />
<br />
神经元及其'''<font color="#ff8000">突触Synapse</font>'''的生理学过程极其复杂，难以进行人工模拟。神经元在所有处理过程中都使用化学模拟信号，这是大脑的一个关键生理学特征。这个特征大大增加了在计算机中复制大脑的难度，因为目前的计算机是完全数字化的。然而，部分特征可以抽象为数学函数，这些函数能够紧密捕捉神经元运行的本质。<br />
<br />
The goal of neuromorphic computing is not to perfectly mimic the brain and all of its functions, but instead to extract what is known of its structure and operations to be used in a practical computing system. No neuromorphic system will claim nor attempt to reproduce every element of neurons and synapses, but all adhere to the idea that computation is highly [[distributed processing|distributed]] throughout a series of small computing elements analogous to a neuron. While this sentiment is standard, researchers chase this goal with different methods.<ref name=":11">{{Cite journal | doi = 10.1088/1741-2560/13/5/051001| title = Large-scale neuromorphic computing systems| journal = Journal of Neural Engineering| volume = 13| pages = 1–15| year = 2016| last1 = Furber | first1 = Steve| issue = 5| pmid = 27529195| bibcode = 2016JNEng..13e1001F| doi-access = free}}</ref><br />
<br />
神经形态计算的目标不是完美地模拟大脑及其所有功能，而是利用已知的大脑结构和运转机制来研发或优化实际的计算系统。任何神经形态学系统都不会声称或试图复制神经元和突触中的每一个元素，但所有人都一致认可将计算高度'''<font color="#ff8000">分散Distribute</font>'''于一系列类似于神经元的小型计算元素的理念。研究人员用不同的方法来追求这一普遍目标。<ref name=":11" /><br />
<br />
<br />
==Examples==<br />
==代表性成果==<br />
As early as 2006, researchers at [[Georgia Tech]] published a field programmable neural array.<ref name=":12">{{Cite book|title = A field programmable neural array|last1 = Farquhar|first1 = Ethan|date = May 2006|journal = IEEE International Symposium on Circuits and Systems|pages = 4114–4117|last2 = Hasler|first2 = Paul.|doi = 10.1109/ISCAS.2006.1693534|isbn = 978-0-7803-9389-9|s2cid = 206966013}}</ref> This chip was the first in a line of increasingly complex arrays of floating gate transistors that allowed programmability of charge on the gates of [[MOSFET]]s to model the channel-ion characteristics of neurons in the brain and was one of the first cases of a silicon programmable array of neurons.<br />
<br />
早在2006年，佐治亚理工学院的研究人员就研发出了一种现场可编程神经阵列。<ref name=":12" />在此之后，一系列越来越复杂的浮栅晶体管阵列被成功研发出来，这些晶体管阵列可以通过在'''<font color="#ff8000">金属-氧化物半导体效应晶体管MOSFET</font>'''的栅极上编程来模拟大脑中神经元的离子通道特性，这也是以硅为主要材料的可编程神经元阵列的最早成功案例之一。<br />
<br />
In November 2011, a group of [[MIT]] researchers created a computer chip that mimics the analog, ion-based communication in a synapse between two neurons using 400 transistors and standard [[CMOS]] manufacturing techniques.<ref name=":13">{{cite web|title=MIT creates "brain chip"|url=http://www.extremetech.com/extreme/105067-mit-creates-brain-chip|access-date=4 December 2012}}</ref><ref name="Neuromorphic silicon paper">{{cite journal|title=Neuromorphic silicon neurons and large-scale neural networks: challenges and opportunities|doi=10.3389/fnins.2011.00108|pmid=21991244|pmc=3181466|volume=5|pages=108|journal=Frontiers in Neuroscience|year=2011|last1=Poon|first1=Chi-Sang|last2=Zhou|first2=Kuan|doi-access=free}}</ref><br />
<br />
2011年11月，麻省理工学院的一组研究人员研发出一种计算机芯片，该芯片上使用标准的'''<font color="#ff8000">互补金属氧化物半导体CMOS</font>'''制造技术集成了400个晶体管，能够模拟神经元间突触中基于离子的通讯。<ref name=":13" /><ref name="Neuromorphic silicon paper" /><br />
<br />
In June 2012, [[spintronic]] researchers at [[Purdue University]] presented a paper on the design of a neuromorphic chip using [[Spin valve|lateral spin valve]]s and [[memristor]]s. They argue that the architecture works similarly to neurons and can therefore be used to test methods of reproducing the brain's processing. In addition, these chips are significantly more energy-efficient than conventional ones.<ref name="Spin Devices Prop">{{Cite arXiv|title=Proposal For Neuromorphic Hardware Using Spin Devices|eprint=1206.3227|last1=Sharad|first1=Mrigank|last2=Augustine|first2=Charles|last3=Panagopoulos|first3=Georgios|last4=Roy|first4=Kaushik|class=cond-mat.dis-nn|year=2012}}</ref><br />
<br />
2012年6月，普渡大学的'''<font color="#ff8000">自旋电子学Spintronic</font>'''研究人员发表了一篇关于利用'''<font color="#ff8000">侧向自旋阀Lateral spin valves</font>'''和'''<font color="#ff8000">记忆电阻器Memristors</font>'''设计神经形态芯片的论文。他们认为，这种芯片结构的工作原理与神经元相似，因此可以用于大脑运行机制的复刻方法的测试。此外，这些芯片在能耗方面明显优于传统芯片。<ref name="Spin Devices Prop" /><br />
<br />
Research at [[HP Labs]] on Mott memristors has shown that while they can be non-[[Volatile memory|volatile]], the volatile behavior exhibited at temperatures significantly below the [[phase transition]] temperature can be exploited to fabricate a [[neuristor]],<ref name=":0" /> a biologically-inspired device that mimics behavior found in neurons.<ref name=":0">{{Cite journal | doi = 10.1038/nmat3510| pmid = 23241533| title = A scalable neuristor built with Mott memristors| journal = Nature Materials| volume = 12| issue = 2| pages = 114–7| year = 2012| last1 = Pickett | first1 = M. D. | last2 = Medeiros-Ribeiro | first2 = G. | last3 = Williams | first3 = R. S. | bibcode = 2013NatMa..12..114P| s2cid = 16271627| url = https://semanticscholar.org/paper/b6ba6f496ace2b947f111059663e76bb60e9efeb}}</ref> In September 2013, they presented models and simulations that show how the spiking behavior of these neuristors can be used to form the components required for a [[Turing machine]].<ref name=":14">{{cite journal|doi=10.1088/0957-4484/24/38/384002|title=Phase transitions enable computational universality in neuristor-based cellular automata|author1=Matthew D Pickett|author2=R Stanley Williams|name-list-style=amp|date=September 2013|publisher=IOP Publishing Ltd|journal=Nanotechnology|volume=24|issue=38|pmid=23999059|bibcode=2013Nanot..24L4002P|at=384002}}</ref><br />
<br />
'''<font color="#ff8000">惠普实验室HP labs</font>'''在莫特记忆电阻器上的研究表明，尽管它们可以是非'''<font color="#ff8000">易失性Volatile</font>'''的，但是在'''<font color="#ff8000">相变Phase transition</font>'''温度以下时表现出的易失性行为可以被用来制造'''<font color="#ff8000">类神经元电阻器Neuristor</font>'''(一种生物学启发的模仿神经元行为的硬件)<ref name=":0" />。2013年9月，他们通过模型和仿真展示了这些类神经元电阻器的脉冲行为如何产生'''<font color="#ff8000">图灵机Turing machine</font>'''的所需元素。<ref name=":14" /><br />
<br />
[[Neurogrid]], built by ''Brains in Silicon'' at [[Stanford University]],<ref name=":15">{{cite journal|last1=Boahen|first1=Kwabena|title=Neurogrid: A Mixed-Analog-Digital Multichip System for Large-Scale Neural Simulations|journal=Proceedings of the IEEE|date=24 April 2014|volume=102|issue=5|pages=699–716|doi=10.1109/JPROC.2014.2313565|s2cid=17176371}}</ref> is an example of hardware designed using neuromorphic engineering principles. The circuit board is composed of 16 custom-designed chips, referred to as NeuroCores. Each NeuroCore's analog circuitry is designed to emulate neural elements for 65536 neurons, maximizing energy efficiency. The emulated neurons are connected using digital circuitry designed to maximize spiking throughput.<ref name=":16">{{cite journal|doi=10.1038/503022a|pmid = 24201264|title = Neuroelectronics: Smart connections|journal = Nature|volume = 503|issue = 7474|pages = 22–4|year = 2013|last1 = Waldrop|first1 = M. Mitchell|bibcode = 2013Natur.503...22W|doi-access = free}}</ref><ref name=":17">{{cite journal|doi=10.1109/JPROC.2014.2313565|title = Neurogrid: A Mixed-Analog-Digital Multichip System for Large-Scale Neural Simulations|journal = Proceedings of the IEEE|volume = 102|issue = 5|pages = 699–716|year = 2014|last1 = Benjamin|first1 = Ben Varkey|last2 = Peiran Gao|last3 = McQuinn|first3 = Emmett|last4 = Choudhary|first4 = Swadesh|last5 = Chandrasekaran|first5 = Anand R.|last6 = Bussat|first6 = Jean-Marie|last7 = Alvarez-Icaza|first7 = Rodrigo|last8 = Arthur|first8 = John V.|last9 = Merolla|first9 = Paul A.|last10 = Boahen|first10 = Kwabena|s2cid = 17176371}}</ref><br />
<br />
'''<font color="#ff8000">神经栅格Neurogrid<ref name=":15" /></font>'''是由斯坦福大学Brains in Silicon公司研发的、使用神经形态工程原理设计的硬件。该电路板由16个定制设计的芯片组成(NeuroCores)。在设计中，每个NeuroCore芯片的模拟电路对65536个神经元的神经元素进行模拟，以最大限度地提高能量效率。模拟出的神经元通过设计的数字电路连接，以最大化脉冲吞吐量。<ref name=":16" /><ref name=":17" /><br />
<br />
A research project with implications for neuromorphic engineering is the [[Human Brain Project]] that is attempting to simulate a complete human brain in a supercomputer using biological data. It is made up of a group of researchers in neuroscience, medicine, and computing.<ref name=":18">{{cite web|title=Involved Organizations|url=http://www.humanbrainproject.eu/partners.html|access-date=22 February 2013|url-status=dead|archive-url=https://web.archive.org/web/20130302142627/http://www.humanbrainproject.eu/partners.html|archive-date=2 March 2013}}</ref> [[Henry Markram]], the project's co-director, has stated that the project proposes to establish a foundation to explore and understand the brain and its diseases, and to use that knowledge to build new computing technologies. The three primary goals of the project are to better understand how the pieces of the brain fit and work together, to understand how to objectively diagnose and treat brain diseases, and to use the understanding of the human brain to develop neuromorphic computers. That the simulation of a complete human brain will require a supercomputer a thousand times more powerful than today's encourages the current focus on neuromorphic computers.<ref name=":19">{{cite web|title=Human Brain Project|url=http://www.humanbrainproject.eu|access-date=22 February 2013}}</ref> $1.3 billion has been allocated to the project by The [[European Commission]].<ref name=":20">{{cite web|title=The Human Brain Project and Recruiting More Cyberwarriors|url=http://www.marketplace.org/topics/tech/human-brain-project-and-recruiting-more-cyberwarriors|access-date=22 February 2013|date=January 29, 2013}}</ref><br />
<br />
'''<font color="#ff8000">人类大脑计划Human Brain Project</font>'''对神经形态工程具有较大影响，其主要任务是尝试用生物数据在超级计算机中模拟完整的人脑。人类大脑计划由神经科学、医学和计算机科学背景的研究人员组成。<ref name=":18" />该项目的联合主管亨利•马克拉姆(Henry Markram)表示，人类大脑计划的目的是建立一个探索和了解脑科学和脑疾病知识的基础，并利用这些知识来构建更先进的计算机技术。这个项目的三个主要目标分别是: 更好地理解大脑的各个部分是如何相互配合协同工作的; 理解如何客观地诊断和治疗脑部疾病; 以及利用对人类大脑的理解来开发神经形态计算机。模拟一个完整的人类大脑需要一台比现在强大一千倍的超级计算机，这不断激发着对神经形态计算机领域的研究兴趣。<ref name=":19" />欧盟委员会已经向人类大脑计划拨款13亿美元。<ref name=":20" /><br />
<br />
Other research with implications for neuromorphic engineering involves the [[BRAIN Initiative]]<ref name="economist">[https://www.economist.com/news/science-and-technology/21582495-computers-will-help-people-understand-brains-better-and-understanding-brains Neuromorphic computing: The machine of a new soul], The Economist, 2013-08-03</ref> and the [[TrueNorth]] chip from [[IBM]].<ref name=":21">{{cite journal|last1=Modha|first1=Dharmendra|title=A million spiking-neuron integrated circuit with a scalable communication network and interface|journal=Science|date=Aug 2014|volume=345|issue=6197|pages=668–673|doi=10.1126/science.1254642|pmid=25104385|bibcode=2014Sci...345..668M|s2cid=12706847}}</ref> Neuromorphic devices have also been demonstrated using nanocrystals, nanowires, and conducting polymers.<ref name=":22">{{Cite web|url=http://jessamynfairfield.com/wp-content/uploads/2017/03/PWMar17Fairfield.pdf|title=Smarter Machines|last=Fairfield|first=Jessamyn|date=March 1, 2017}}</ref><br />
<br />
其他与神经形态工程有关的研究还包括'''<font color="#ff8000">脑计划BRAIN initiative</font>'''，<ref name="economist" />和IBM研发的'''<font color="#ff8000">TrueNorth</font>'''芯片。<ref name=":21" />使用纳米晶体、纳米线和导电聚合物也能够用于制造神经形态学硬件。<ref name=":22" /><br />
<br />
[[Intel]] unveiled its neuromorphic research chip, called “[[Intel Loihi|Loihi]]”, in October 2017. The chip uses an asynchronous [[spiking neural network]] (SNN) to implement adaptive self-modifying event-driven fine-grained parallel computations used to implement learning and inference with high efficiency.<ref name=":23">{{cite journal |last1=Davies |first1=Mike |title=Loihi: A Neuromorphic Manycore Processor with On-Chip Learning |journal=IEEE Micro |date=January 16, 2018 |volume=38 |issue=1 |pages=82–99 |display-authors=etal|doi=10.1109/MM.2018.112130359 |s2cid=3608458 }}</ref><ref name="Morris2017">{{cite web |last1=Morris |first1=John |title=Why Intel built a neuromorphic chip |url=https://www.zdnet.com/article/why-intel-built-a-neuromorphic-chip/ |website=ZDNet |access-date=17 August 2018 |language=en}}</ref><br />
<br />
2017年10月，英特尔发布了神经形态芯片'''<font color="#ff8000">Loihi</font>'''。该芯片采用异步'''<font color="#ff8000">脉冲神经网络Spiking neural network</font>'''实现了自适应、自修改、事件驱动的细粒度并行计算，实现了高效的学习和推理。<ref name=":23" /><ref name="Morris2017" /><br />
<br />
[[IMEC]], a Belgium-based nanoelectronics research center, demonstrated the world's first self-learning neuromorphic chip. The brain-inspired chip, based on OxRAM technology, has the capability of self-learning and has been demonstrated to have the ability to compose music.<ref name=":24">{{cite web |title=Imec demonstrates self-learning neuromorphic chip that composes music |url=https://www.imec-int.com/en/articles/imec-demonstrates-self-learning-neuromorphic-chip-that-composes-music |website=IMEC International |access-date=1 October 2019}}</ref> IMEC released the 30-second tune composed by the prototype. The chip was sequentially loaded with songs in the same time signature and style. The songs were old Belgian and French flute minuets, from which the chip learned the rules at play and then applied them.<ref name=":25">{{cite web|last1=Bourzac|first1=Katherine|title=A Neuromorphic Chip That Makes Music|url=https://spectrum.ieee.org/a-neuromorphic-chip-that-makes-music|url-status=live|access-date=1 October 2019|website=IEEE Spectrum|date=May 23, 2017}}</ref><br />
<br />
比利时的'''<font color="#ff8000">微电子研究中心IMEC</font>'''研发了世界上首个自学习神经形态芯片。这种基于'''<font color="#32CD32"> OxRAM(filamentary-oxide-based resistive memory technology)</font>''' 技术的大脑启发芯片具有自学习能力，并且已被证明具有创作音乐的能力。<ref name=":24" />IMEC发布了由芯片原型机谱写的30秒旋律。向芯片加载一系列特征、风格统一的歌曲（古代比利时和法国长笛小步舞曲），芯片从中学习相关规则并将其应用于创作。<ref name=":25" /><br />
<br />
[[Blue Brain Project|The Blue Brain Project]], led by Henry Markram, aims to build biologically detailed digital reconstructions and simulations of the mouse brain. The Blue Brain Project has created in silico models of rodent brains, while attempting to replicate as many details about its biology as possible. The supercomputer-based simulations offer new perspectives on understanding the structure and functions of the brain.<br />
<br />
由亨利·马克拉姆领导的'''<font color="#ff8000">蓝脑计划The Blue Brain Project</font>'''旨在建立小鼠大脑生理学细节的数字重建和模拟。蓝脑计划已经建立了啮齿动物大脑的电子模型，同时进行着尽可能多地复制其生理学细节的尝试。基于超级计算机的模拟为理解大脑的结构和功能提供了新的视角。<br />
<br />
The European Union funded a series of projects at the University of Heidelberg, which led to the development of <br />
[[BrainScaleS]] (brain-inspired multiscale computation in neuromorphic hybrid systems), a hybrid analog [[neuromorphic]] supercomputer located at Heidelberg University, Germany. It was developed as part of the [[Human Brain Project]] neuromorphic computing platform and is the complement to the [[SpiNNaker]] supercomputer (which is based on digital technology). The architecture used in BrainScaleS mimics biological neurons and their connections on a physical level; additionally, since the components are made of silicon, these model neurons operate on average 864 times (24 hours of real time is 100 seconds in the machine simulation) that of their biological counterparts.<ref name=":26">{{Cite web|date=2016-03-21|title=Beyond von Neumann, Neuromorphic Computing Steadily Advances|url=https://www.hpcwire.com/2016/03/21/lacking-breakthrough-neuromorphic-computing-steadily-advance/|access-date=2021-10-08|website=HPCwire|language=en-US}}</ref><br />
<br />
欧盟资助了海德堡大学一系列促进BrainScaleS(神经形态混合系统中受大脑启发的多尺度计算)发展的项目，这是一台位于德国海德堡大学的混合模拟'''<font color="#ff8000">神经形态Neuromorphic</font>'''超级计算机。它是作为人类大脑计划中神经形态计算平台的一部分而开发的，是'''<font color="#ff8000">SpiNNaker</font>'''超级计算机(基于数字技术)的补充。BrainScaleS中使用的体系架构模拟了生物神经元及其在物理层面上的连接；此外，由于这些组件是由硅制成的，这些模型神经元平均运行速度是生物神经元的864倍，这意味着在机器模拟中，24小时的实时时间仅为100秒。<ref name=":26" /><br />
<br />
===Neuromorphic sensors===<br />
===神经形态传感器===<br />
The concept of neuromorphic systems can be extended to sensors (not just to computation). An example of this applied to detecting [[light]] is the [[retinomorphic sensor]] or, when employed in an array, the [[event camera]].<br />
<br />
神经形态系统的概念可以扩展到传感器(而不仅仅是计算单元)。一个用于检测光线的例子是'''<font color="#ff8000">类视网膜传感器Retinomorphic sensor</font>'''，或者'''<font color="#ff8000">事件摄像机Event camera</font>阵列。'''<br />
<br />
<br />
==Ethical considerations ==<br />
==伦理问题==<br />
While the interdisciplinary concept of neuromorphic engineering is relatively new, many of the same ethical considerations apply to neuromorphic systems as apply to [[human-like machines]] and [[artificial intelligence]] in general. However, the fact that neuromorphic systems are designed to mimic a [[human brain]] gives rise to unique ethical questions surrounding their usage.<br />
<br />
虽然神经形态工程这个跨学科概念相对较新，但许多适用于'''<font color="#ff8000">类人机器Human-like machines</font>'''和人工智能的伦理考虑大体上也适用于神经形态系统。另一方面，神经形态系统是为了模仿人类大脑而设计，这一事实引起了一些针对性的独特伦理问题。<br />
<br />
However, the practical debate is that neuromorphic hardware as well as artificial "neural networks" are immensely simplified models of how the brain operates or processes information at a much lower [[complex system|complexity in terms of size and functional technology]] and a much more regular structure in terms of [[brain connectivity|connectivity]]. Comparing [[neuromorphic chip]]s to the brain is a very crude comparison similar to comparing a plane to a bird just because they both have wings and a tail. The fact is that neural cognitive systems are many orders of magnitude more [[Energy efficiency (physics)|energy-]] and compute-efficient than current state-of-the-art AI and neuromorphic engineering is an attempt to narrow this gap by inspiring from the brain's mechanism just like many engineering designs have [[bioengineering|bio-inspired features]].<br />
<br />
然而，这种争论的实际情况是，神经形态硬件和人工“神经网络”是对大脑运作或处理信息过程的极其简化的模型，在大小和功能技术方面的'''<font color="#ff8000">复杂性Complexity</font>'''相比而言比较低，在'''<font color="#ff8000">连接性Connectivity</font>'''方面也具有更加规则的结构。将'''<font color="#ff8000">神经形态芯片Neuromorphic chips</font>'''与大脑进行比较是一种非常粗糙的比较，类似于仅仅因为飞机有翅膀和尾巴就将它与鸟进行比较。事实上，当前最先进的人工智能在'''<font color="#ff8000">能耗效率Energy-efficiency</font>'''和计算效率方面距离人脑神经认知系统仍有较大差距，而神经形态工程只是一种通过从大脑机制中获得灵感来缩小这种差距的尝试，就像许多工程设计中都具有'''<font color="#ff8000">生物启发的特征Bio-inspired features</font>'''一样。<br />
<br />
===Democratic concerns===<br />
===公众担忧===<br />
Significant ethical limitations may be placed on neuromorphic engineering due to public perception.<ref name=":27">{{Cite report|url=https://ai100.stanford.edu/sites/g/files/sbiybj9861/f/ai_100_report_0831fnl.pdf|title=Artificial Intelligence and Life in 2030|author=2015 Study Panel|date=September 2016|work=One Hundred Year Study on Artificial Intelligence (AI100)|publisher=Stanford University}}</ref> Special [[Eurobarometer]] 382: Public Attitudes Towards Robots, a survey conducted by the European Commission, found that 60% of [[European Union]] citizens wanted a ban of robots in the care of children, the elderly, or the disabled. Furthermore, 34% were in favor of a ban on robots in education, 27% in healthcare, and 20% in leisure. The European Commission classifies these areas as notably “human.” The report cites increased public concern with robots that are able to mimic or replicate human functions. Neuromorphic engineering, by definition, is designed to replicate the function of the human brain.<ref name=":1">{{Cite web|url=http://ec.europa.eu/commfrontoffice/publicopinion/archives/ebs/ebs_382_en.pdf|title=Special Eurobarometer 382: Public Attitudes Towards Robots|last=European Commission|date=September 2012|website=European Commission}}</ref><br />
<br />
由于公众认知的相关忧虑，神经形态工程学可能会受到严重的伦理限制。<ref name=":27" />欧盟委员会进行的一项调查发现，60% 的欧盟公民希望禁止机器人参与照顾儿童、老人或残疾人的工作。此外，34% 的人支持禁止机器人用于教育，27% 的人支持禁止机器人用于医疗保健，20% 的人支持禁止机器人用于休闲。欧盟委员会将以上领域划入“人类”范畴。报告指出，公众越来越关注能够模仿或复制人类功能的机器人。神经形态工程，顾名思义，是为了复制人脑的功能而设计的。<ref name=":1" /><br />
<br />
The democratic concerns surrounding neuromorphic engineering are likely to become even more profound in the future. The European Commission found that EU citizens between the ages of 15 and 24 are more likely to think of robots as human-like (as opposed to instrument-like) than EU citizens over the age of 55. When presented an image of a robot that had been defined as human-like, 75% of EU citizens aged 15–24 said it corresponded with the idea they had of robots while only 57% of EU citizens over the age of 55 responded the same way. The human-like nature of neuromorphic systems, therefore, could place them in the categories of robots many EU citizens would like to see banned in the future.<ref name=":1" /><br />
<br />
围绕神经形态工程的公众担忧可能在未来变得更加深刻。欧盟委员会发现，相比于55岁以上的欧盟公民，15至24岁的欧盟公民更有可能认为机器人像人(而不是像仪器)。当看到一张“类人”机器人的图片时，年龄在15岁至24岁之间的欧盟公民中有75% 的人表示这符合他们对机器人的想法，而55岁以上的欧盟公民中只有57% 的人有同样的反应。因此，神经形态系统可能因为其类似人类的特性而被归入许多欧盟公民希望在未来禁止使用的机器人类别。<ref name=":1" /><br />
<br />
=== Personhood===<br />
=== 人格问题===<br />
As neuromorphic systems have become increasingly advanced, some scholars{{who|date=August 2021}} have advocated for granting [[personhood]] rights to these systems. If the brain is what grants humans their personhood, to what extent does a neuromorphic system have to mimic the human brain to be granted personhood rights? Critics of technology development in the [[Human Brain Project]], which aims to advance brain-inspired computing, have argued that advancement in neuromorphic computing could lead to machine consciousness or personhood.<ref name=":28">{{Cite journal|last=Aicardi|first=Christine|date=September 2018|title=Accompanying technology development in the Human Brain Project: From foresight to ethics management|journal=Futures|volume=102|pages=114–124|doi=10.1016/j.futures.2018.01.005|doi-access=free}}</ref> If these systems are to be treated as people, critics argue, then many tasks humans perform using neuromorphic systems, including the act of termination of neuromorphic systems, may be morally impermissible as these acts would violate the autonomy of the neuromorphic systems.<ref name=":29">{{Cite journal|last=Lim|first=Daniel|date=2014-06-01|title=Brain simulation and personhood: a concern with the Human Brain Project|journal=Ethics and Information Technology|language=en|volume=16|issue=2|pages=77–89|doi=10.1007/s10676-013-9330-5|s2cid=17415814|issn=1572-8439}}</ref><br />
<br />
随着神经形态系统的日益发展，一些学者主张赋予这些系统'''<font color="ff8000">人格Personhood</font>'''权。如果是大脑赋予了人类人格，那么在多大程度上模仿人类大脑的神经形态系统才能被赋予人格权利？“人类大脑计划”旨在推进以大脑为灵感的计算机技术发展，该计划的批评者认为，神经形态计算机技术的进步可能导致机器意识或人格的形成。<ref name=":28" />这些批评者认为，如果这些系统被当作人来对待，那么人类使用神经形态系统执行任务（包括终止神经形态系统）的行为，在道德上就可能是不被允许的，因为这些行为将违反神经形态系统的自主性。<ref name=":29" /><br />
<br />
<br />
==Dual use (military applications)==<br />
==军民两用技术==<br />
The [[Joint Artificial Intelligence Center]], a branch of the U.S. military, is a center dedicated to the procurement and implementation of AI software and neuromorphic hardware for combat use. Specific applications include smart headsets/goggles and robots. JAIC intends to rely heavily on neuromorphic technology to connect "every fighter every shooter" within a network of neuromorphic-enabled units.<br />
<br />
The Joint Artificial Intelligence Center, a branch of the U.S. military, is a center dedicated to the procurement and implementation of AI software and neuromorphic hardware for combat use. Specific applications include smart headsets/goggles and robots. JAIC intends to rely heavily on neuromorphic technology to connect "every fighter every shooter" within a network of neuromorphic-enabled units.<br />
<br />
军民两用联合人工智能中心是美国军队的一个分支，专门从事采购和实施用于战斗的人工智能软件和神经形态硬件。具体应用包括智能耳机/护目镜和机器人。JAIC 打算严重依赖神经形态技术来连接神经形态单位网络中的“每个战士每个射手”。<br />
<br />
==Legal considerations==<br />
== 法律问题==<br />
Skeptics have argued that there is no way to apply the electronic personhood, the concept of personhood that would apply to neuromorphic technology, legally. In a letter signed by 285 experts in law, robotics, medicine, and ethics opposing a European Commission proposal to recognize “smart robots” as legal persons, the authors write, “A legal status for a robot can’t derive from the [[Natural person|Natural Person]] model, since the robot would then hold [[human rights]], such as the right to dignity, the right to its integrity, the right to remuneration or the right to citizenship, thus directly confronting the Human rights. This would be in contradiction with the [[Charter of Fundamental Rights of the European Union]] and the [[Convention for the Protection of Human Rights and Fundamental Freedoms]].”<ref>{{Cite web|url=http://www.robotics-openletter.eu/|title=Robotics Openletter {{!}} Open letter to the European Commission|language=fr-FR|access-date=2019-05-10}}</ref><br />
<br />
Skeptics have argued that there is no way to apply the electronic personhood, the concept of personhood that would apply to neuromorphic technology, legally. In a letter signed by 285 experts in law, robotics, medicine, and ethics opposing a European Commission proposal to recognize “smart robots” as legal persons, the authors write, “A legal status for a robot can’t derive from the Natural Person model, since the robot would then hold human rights, such as the right to dignity, the right to its integrity, the right to remuneration or the right to citizenship, thus directly confronting the Human rights. This would be in contradiction with the Charter of Fundamental Rights of the European Union and the Convention for the Protection of Human Rights and Fundamental Freedoms.”<br />
<br />
怀疑论者认为，在法律上没有办法应用电子人格，这个人格概念将适用于神经形态技术。在一封由285名法律、机器人、医学和伦理学专家签名的信中，作者们反对欧盟委员会承认“智能机器人”为法人的提议，他们写道，“机器人的法律地位不能从自然人模型中推导出来，因为机器人将拥有人权，如尊严权、完整权、报酬权或公民权，从而直接面对人权。这将有悖于《欧洲联盟基本权利宪章和《保护人权和基本自由公约》。”<br />
<br />
===Ownership and property rights===<br />
===所有权及财产权问题 ===<br />
There is significant legal debate around property rights and artificial intelligence. In ''Acohs Pty Ltd v. Ucorp Pty Ltd'', Justice Christopher Jessup of the [[Federal Court of Australia]] found that the [[source code]] for [[Material safety data sheets|Material Safety Data Sheets]] could not be [[Copyright law of Australia|copyrighted]] as it was generated by a [[software interface]] rather than a human author.<ref>{{Cite web|url=http://www.lavan.com.au/advice/intellectual_property_technology/copyright_in_source_code_and_digital_products|title=Copyright in source code and digital products|last=Lavan|website=Lavan|language=en|access-date=2019-05-10}}</ref> The same question may apply to neuromorphic systems: if a neuromorphic system successfully mimics a human brain and produces a piece of original work, who, if anyone, should be able to claim ownership of the work?<ref>{{cite journal |last1=Eshraghian|first1=Jason K. |title=Human Ownership of Artificial Creativity |journal=Nature Machine Intelligence |date=9 March 2020 |volume=2 |pages=157–160 |doi=10.1038/s42256-020-0161-x}}</ref><br />
<br />
There is significant legal debate around property rights and artificial intelligence. In Acohs Pty Ltd v. Ucorp Pty Ltd, Justice Christopher Jessup of the Federal Court of Australia found that the source code for Material Safety Data Sheets could not be copyrighted as it was generated by a software interface rather than a human author. The same question may apply to neuromorphic systems: if a neuromorphic system successfully mimics a human brain and produces a piece of original work, who, if anyone, should be able to claim ownership of the work?<br />
<br />
围绕财产权和人工智能有着重大的法律争论。在 Acohs Pty Ltd 诉 Ucorp Pty Ltd 一案中，澳大利亚联邦法院的克里斯托弗 · 杰瑟普法官发现，材料安全数据表的源代码不能受版权保护，因为它是由软件界面而不是人工作者生成的。同样的问题可能也适用于神经形态系统: 如果一个神经形态系统成功地模仿了人类的大脑并产生了一部原创作品，那么谁，如果有人，应该声称拥有这部作品的所有权？<br />
<br />
==Neuromemristive systems==<br />
==神经记忆电阻系统==<br />
Neuromemristive systems are a subclass of neuromorphic computing systems that focus on the use of [[memristors]] to implement [[neuroplasticity]]. While neuromorphic engineering focuses on mimicking biological behavior, neuromemristive systems focus on abstraction.<ref>{{Cite web|url=https://digitalops.sandia.gov/Mediasite/Play/a10cf6ceb55d47608bb8326dd00e46611d|title=002.08 N.I.C.E. Workshop 2014: Towards Intelligent Computing with Neuromemristive Circuits and Systems - Feb. 2014|website=digitalops.sandia.gov|access-date=2019-08-26}}</ref> For example, a neuromemristive system may replace the details of a [[Cerebral cortex|cortical]] microcircuit's behavior with an abstract neural network model.<ref>C. Merkel and D. Kudithipudi, "Neuromemristive extreme learning machines for pattern classification," ISVLSI, 2014.</ref><br />
<br />
Neuromemristive systems are a subclass of neuromorphic computing systems that focus on the use of memristors to implement neuroplasticity. While neuromorphic engineering focuses on mimicking biological behavior, neuromemristive systems focus on abstraction. For example, a neuromemristive system may replace the details of a cortical microcircuit's behavior with an abstract neural network model.C. Merkel and D. Kudithipudi, "Neuromemristive extreme learning machines for pattern classification," ISVLSI, 2014.<br />
<br />
神经记忆电阻系统是神经形态计算系统的一个亚类，主要研究利用记忆电阻器实现神经可塑性。神经形态工程的重点是模拟生物行为，而神经记忆电阻系统的重点是提取。例如，一个神经记忆系统可以用一个抽象的神经网络模型取代皮层微电路的行为细节。默克尔和 d. Kudithipudi，“用于模式分类的神经记忆极端学习机器”，ISVLSI，2014。<br />
<br />
There exist several neuron inspired threshold logic functions<ref name="Maan 1–13" /> implemented with memristors that have applications in high level [[pattern recognition]] applications. Some of the applications reported recently include [[speech recognition]],<ref>{{Cite journal|title = Memristor pattern recogniser: isolated speech word recognition|journal = Electronics Letters|pages = 1370–1372|volume = 51|issue = 17|doi = 10.1049/el.2015.1428|first1 = A.K.|last1 = Maan|first2 = A.P.|last2 = James|first3 = S.|last3 = Dimitrijev|year = 2015|bibcode = 2015ElL....51.1370M|hdl = 10072/140989|s2cid = 61454815|url = https://semanticscholar.org/paper/48d3ab11ec6e213b62f11eedcfb7b7febb058674|hdl-access = free}}</ref> [[face recognition]]<ref>{{Cite journal|title = Memristive Threshold Logic Face Recognition|journal = Procedia Computer Science|date = 2014-01-01|pages = 98–103|volume = 41|series = 5th Annual International Conference on Biologically Inspired Cognitive Architectures, 2014 BICA|doi = 10.1016/j.procs.2014.11.090|first1 = Akshay Kumar|last1 = Maan|first2 = Dinesh S.|last2 = Kumar|first3 = Alex Pappachen|last3 = James|doi-access = free}}</ref> and [[object recognition]].<ref>{{Cite journal|title = Memristive Threshold Logic Circuit Design of Fast Moving Object Detection|journal = IEEE Transactions on Very Large Scale Integration (VLSI) Systems|date = 2015-10-01|issn = 1063-8210|pages = 2337–2341|volume = 23|issue = 10|doi = 10.1109/TVLSI.2014.2359801|first1 = A.K.|last1 = Maan|first2 = D.S.|last2 = Kumar|first3 = S.|last3 = Sugathan|first4 = A.P.|last4 = James|arxiv = 1410.1267|s2cid = 9647290}}</ref> They also find applications in replacing conventional digital logic gates.<ref>{{Cite journal|title = Resistive Threshold Logic|journal = IEEE Transactions on Very Large Scale Integration (VLSI) Systems|date = 2014-01-01|issn = 1063-8210|pages = 190–195|volume = 22|issue = 1|doi = 10.1109/TVLSI.2012.2232946|first1 = A.P.|last1 = James|first2 = L.R.V.J.|last2 = Francis|first3 = D.S.|last3 = Kumar|arxiv = 1308.0090|s2cid = 7357110}}</ref><ref>{{Cite journal|title = Threshold Logic Computing: Memristive-CMOS Circuits for Fast Fourier Transform and Vedic Multiplication|journal = IEEE Transactions on Very Large Scale Integration (VLSI) Systems|date = 2015-11-01|issn = 1063-8210|pages = 2690–2694|volume = 23|issue = 11|doi = 10.1109/TVLSI.2014.2371857|first1 = A.P.|last1 = James|first2 = D.S.|last2 = Kumar|first3 = A.|last3 = Ajayan|arxiv = 1411.5255|s2cid = 6076956}}</ref><br />
<br />
There exist several neuron inspired threshold logic functions implemented with memristors that have applications in high level pattern recognition applications. Some of the applications reported recently include speech recognition, face recognition and object recognition. They also find applications in replacing conventional digital logic gates.<br />
<br />
利用记忆电阻器实现的神经元启发阈值逻辑函数在高级模式识别中有着广泛的应用。最近报道的一些应用包括语音识别、人脸识别和物体识别。它们还可以用来取代传统的数字逻辑门。<br />
<br />
For ideal passive memristive circuits there is an exact equation (Caravelli-Traversa-[[Di Ventra]] equation) for the internal memory of the circuit:<ref>{{cite journal |last=Caravelli |display-authors=etal|arxiv=1608.08651 |title=The complex dynamics of memristive circuits: analytical results and universal slow relaxation |year=2017 |doi=10.1103/PhysRevE.95.022140 |pmid= 28297937 |volume=95 |issue= 2 |pages= 022140 |journal=Physical Review E|bibcode=2017PhRvE..95b2140C |s2cid=6758362}}</ref><br />
<br />
For ideal passive memristive circuits there is an exact equation (Caravelli-Traversa-Di Ventra equation) for the internal memory of the circuit:<br />
<br />
对于理想的无源记忆电路，电路的内部记忆有一个精确的方程(Caravelli-Traversa-Di Ventra 方程) :<br />
<br />
:<math> \frac{d}{dt} \vec{W} = \alpha \vec{W}-\frac{1}{\beta} (I+\xi \Omega W)^{-1} \Omega \vec S </math><br />
<br />
:\frac{d}{dt} \vec{W} = \alpha \vec{W}-\frac{1}{\beta} (I+\xi \Omega W)^{-1} \Omega \vec S<br />
<br />
:\frac{d}{dt} \vec{W} = \alpha \vec{W}-\frac{1}{\beta} (I+\xi \Omega W)^{-1} \Omega \vec S<br />
<br />
as a function of the properties of the physical memristive network and the external sources. In the equation above, <math>\alpha</math> is the "forgetting" time scale constant, <math> \xi=r-1</math> and <math>r=\frac{R_\text{off}}{R_\text{on}}</math> is the ratio of ''off'' and ''on'' values of the limit resistances of the memristors, <math> \vec S </math> is the vector of the sources of the circuit and <math>\Omega</math> is a projector on the fundamental loops of the circuit. The constant <math>\beta</math> has the dimension of a voltage and is associated to the properties of the [[memristor]]; its physical origin is the charge mobility in the conductor. The diagonal matrix and vector <math>W=\operatorname{diag}(\vec W)</math> and <math>\vec W</math> respectively, are instead the internal value of the memristors, with values between 0 and 1. This equation thus requires adding extra constraints on the memory values in order to be reliable.<br />
<br />
<nowiki>as a function of the properties of the physical memristive network and the external sources. In the equation above, \alpha is the "forgetting" time scale constant, \xi=r-1 and r=\frac{R_\text{off}}{R_\text{on}} is the ratio of off and on values of the limit resistances of the memristors, \vec S is the vector of the sources of the circuit and \Omega is a projector on the fundamental loops of the circuit. The constant \beta has the dimension of a voltage and is associated to the properties of the memristor; its physical origin is the charge mobility in the conductor. The diagonal matrix and vector W=\operatorname{diag}(\vec W) and \vec W respectively, are instead the internal value of the memristors, with values between 0 and 1. This equation thus requires adding extra constraints on the memory values in order to be reliable.</nowiki><br />
<br />
<nowiki>作为物理记忆网络和外部源的性质的函数。在上述方程中，α 是“遗忘”时间尺度常数，xi = r-1，r = frac { r _ text { off }{ on }{ r _ text { on }}是记忆电阻器极限电阻的开关和开关值之比，vec s 是电路源的矢量，Omega 是电路基本环路的投影仪。常数 β 具有电压的尺寸，与记忆电阻器的特性有关; 它的物理起源是导体中的电荷迁移率。对角矩阵和向量 w = 操作者名{ diag }(vec w)和 vec w 分别是记忆电阻器的内值，值在0到1之间。因此，这个等式需要在内存值上添加额外的约束，以保证可靠性。</nowiki><br />
<br />
==See also==<br />
==相关词条==<br />
{{Columns-list|colwidth=18em|<br />
* [[AI accelerator (computer hardware)|AI accelerator]]<br />
* [[Artificial brain]]<br />
* [[Biomorphic]]<br />
* [[Cognitive computer]]<br />
* [[Computation and Neural Systems]]<br />
* [[Differentiable programming]]<br />
* [[Event camera]]<br />
* [[Neurorobotics]]<br />
* [[Optical flow sensor]]<br />
* [[Physical neural network]]<br />
* [[SpiNNaker]]<br />
* [[SyNAPSE]]<br />
* [[Retinomorphic sensor]]<br />
* [[Vision chip]]<br />
* [[Vision processing unit]]<br />
* [[Zeroth (software)]]<br />
* [[Hardware for artificial intelligence]]<br />
}}<br />
<br />
{{Portal bar|Electronics}}<br />
<br />
==References==<br />
{{Reflist|40em}}<br />
<br />
==External links==<br />
<!--======================== {{No more links}} ============================<br />
| PLEASE BE CAUTIOUS IN ADDING MORE LINKS TO THIS ARTICLE. Wikipedia |<br />
| is not a collection of links nor should it be used for advertising. |<br />
| |<br />
| Excessive or inappropriate links WILL BE DELETED. |<br />
| See [[Wikipedia:External links]] & [[Wikipedia:Spam]] for details. |<br />
| |<br />
| If there are already plentiful links, please propose additions or |<br />
| replacements on this article's discussion page, or submit your link |<br />
| to the relevant category at the Open Directory Project (dmoz.org) |<br />
| and link back to that category using the {{dmoz}} template. |<br />
======================={{No more links}}=============================--><br />
{{Commons category|Neuromorphic Engineering}}<br />
*[https://web.archive.org/web/20150727034331/http://ine-web.org/workshops/workshops-overview Telluride Neuromorphic Engineering Workshop]<br />
*[https://archive.today/20130115190057/http://capocaccia.ethz.ch/ CapoCaccia Cognitive Neuromorphic Engineering Workshop]<br />
*[https://web.archive.org/web/20190716132350/http://www.ine-web.org/ Institute of Neuromorphic Engineering]<br />
*[http://www.ine-news.org/ INE news site].<br />
*[http://www.frontiersin.org/neuromorphic_engineering Frontiers in Neuromorphic Engineering Journal]<br />
*[http://www.cns.caltech.edu/ Computation and Neural Systems] department at the [[California Institute of Technology]].<br />
*[http://www.humanbrainproject.eu/ Human Brain Project official site]<br />
*[https://www.the-scientist.com/features/building-a-silicon-brain-65738 Building a Silicon Brain:] Computer chips based on biological neurons may help simulate larger and more-complex brain models. May 1, 2019. SANDEEP RAVINDRAN<br />
<br />
<br />
*Telluride Neuromorphic Engineering Workshop<br />
*CapoCaccia Cognitive Neuromorphic Engineering Workshop<br />
*Institute of Neuromorphic Engineering<br />
*INE news site.<br />
*Frontiers in Neuromorphic Engineering Journal<br />
* Computation and Neural Systems department at the California Institute of Technology.<br />
*Human Brain Project official site<br />
*Building a Silicon Brain: Computer chips based on biological neurons may help simulate larger and more-complex brain models. May 1, 2019. SANDEEP RAVINDRAN<br />
<br />
=<nowiki>外部链接</nowiki>= <br />
*碲化物神经形态工程工作室<br />
*CapoCaccia 认知神经形态工程工作室<br />
*神经形态工程研究所<br />
*INE 新闻站点。<br />
*《神经形态工程学前沿》<br />
*加州理工学院计算与神经系统系。<br />
*人脑项目官方网站<br />
*打造硅脑: 基于生物神经元的计算机芯片可能有助于模拟更大、更复杂的大脑模型。2019年5月1日。SANDEEP RAVINDRAN<br />
<br />
{{Differentiable computing}}<br />
{{Authority control}}<br />
<br />
[[Category:Electrical engineering]]<br />
[[Category:Neuroscience]]<br />
[[分类:AI accelerators|*]]<br />
[[Category:Artificial intelligence]]<br />
[[Category:Robotics]]<br />
<br />
Category:Electrical engineering<br />
Category:Neuroscience<br />
*<br />
Category:Artificial intelligence<br />
Category:Robotics<br />
<br />
类别: 电气工程类别: 神经科学 * 类别: 人工智能类别: 机器人<br />
<br />
<noinclude><br />
<br />
<small>This page was moved from [[wikipedia:en:Neuromorphic engineering]]. Its edit history can be viewed at [[神经形态计算/edithistory]]</small></noinclude><br />
<br />
[[Category:待整理页面]]<br />
<br />
<references /></div>Shenky20https://wiki.swarma.org/index.php?title=%E7%A5%9E%E7%BB%8F%E5%BD%A2%E6%80%81%E8%AE%A1%E7%AE%97&diff=30745神经形态计算2022-04-28T09:37:38Z<p>Shenky20：/* 伦理问题 */</p>
<hr />
<div><br />
此词条由神经动力学模型读书会词条梳理志愿者Shenky20翻译审校，未经专家审校，带来阅读不便，请见谅。<br />
<br />
{{Use American English|date = January 2019}}<br />
{{Use mdy dates|date = January 2019}}<br />
<br />
'''Neuromorphic engineering''', also known as '''neuromorphic computing''',<ref name=":3">{{Cite journal | doi = 10.1145/2601069| title = Neuromorphic computing gets ready for the (really) big time| journal = [[Communications of the ACM]]| volume = 57| issue = 6| year = 2014| last1 = Monroe | first1 = D. | pages = 13–15| s2cid = 20051102}}</ref><ref name=":4">{{Cite journal | doi = 10.1088/0957-4484/21/17/175202| pmid = 20368686| title = Nanotube devices based crossbar architecture: Toward neuromorphic computing| journal = Nanotechnology| volume = 21| issue = 17| pages = 175202| year = 2010| last1 = Zhao | first1 = W. S. | last2 = Agnus | first2 = G. | last3 = Derycke | first3 = V. | last4 = Filoramo | first4 = A. | last5 = Bourgoin | first5 = J. -P. | last6 = Gamrat | first6 = C. | bibcode = 2010Nanot..21q5202Z| url = https://zenodo.org/record/3428659}}</ref><ref name="humanbrainproject">{{YouTube|id=6RoiZ90mGfw|title=The Human Brain Project SP 9: Neuromorphic Computing Platform}}</ref> is the use of [[very-large-scale integration]] (VLSI) systems containing electronic [[analog circuit]]s to mimic neuro-biological architectures present in the nervous system. A neuromorphic computer/chip is any device that uses physical artificial neurons (made from silicon) to do computations.<ref name=":5">{{cite journal|last1=Mead|first1=Carver|title=Neuromorphic electronic systems|journal=Proceedings of the IEEE|date=1990|volume=78|issue=10|pages=1629–1636|doi=10.1109/5.58356|url=https://authors.library.caltech.edu/53090/1/00058356.pdf}}</ref><ref name=":2" /> In recent times, the term ''neuromorphic'' has been used to describe analog, digital, mixed-mode analog/digital VLSI, and software systems that implement models of [[neural system]]s (for [[perception]], [[motor control]], or [[multisensory integration]]). The implementation of neuromorphic computing on the hardware level can be realized by oxide-based [[memristor]]s,<ref name="Maan 1–13">{{Cite journal|last1=Maan|first1=A. K.|last2=Jayadevi|first2=D. A.|last3=James|first3=A. P.|date=2016-01-01|title=A Survey of Memristive Threshold Logic Circuits|journal=IEEE Transactions on Neural Networks and Learning Systems|volume=PP|issue=99|pages=1734–1746|doi=10.1109/TNNLS.2016.2547842|pmid=27164608|issn=2162-237X|arxiv=1604.07121|bibcode=2016arXiv160407121M|s2cid=1798273}}</ref> spintronic memories, threshold switches, and [[transistor]]s.<ref name=":6">{{Cite journal|title = Mott Memory and Neuromorphic Devices|journal = Proceedings of the IEEE|date = 2015-08-01|issn = 0018-9219|pages = 1289–1310|volume = 103|issue = 8|doi = 10.1109/JPROC.2015.2431914|first1 = You|last1 = Zhou|first2 = S.|last2 = Ramanathan|s2cid = 11347598|url = https://zenodo.org/record/895565}}</ref><ref name=":2">{{Cite document|title=Neuromorphic Circuits With Neural Modulation Enhancing the Information Content of Neural Signaling {{!}} International Conference on Neuromorphic Systems 2020|language=EN|doi=10.1145/3407197.3407204|s2cid=220794387}}</ref> Training software-based neuromorphic systems of spiking neural networks can be achieved using error backpropagation, e.g., using [[Python (programming language)|Python]] based frameworks such as snnTorch,<ref name=":7">{{cite journal |last1=Eshraghian|first1=Jason K.|last2=Ward|first2=Max|last3=Neftci |first3=Emre|last4=Wang|first4=Xinxin|last5=Lenz|first5=Gregor|last6=Dwivedi|first6=Girish|last7=Bennamoun|first7=Mohammed|last8=Jeong|first8=Doo Seok|last9=Lu|first9=Wei D.|title=Training Spiking Neural Networks Using Lessons from Deep Learning |date=1 October 2021 |arxiv=2109.12894 }}</ref> or using canonical learning rules from the biological learning literature, e.g., using BindsNet.<ref name=":8">{{Cite web | url=https://github.com/Hananel-Hazan/bindsnet | title=Hananel-Hazan/bindsnet: Simulation of spiking neural networks (SNNs) using PyTorch.| date=31 March 2020}}</ref><br />
<br />
<br />
'''<font color="#ff8000">神经形态工程Neuromorphic engineering</font>'''(也称为'''<font color="#ff8000">神经形态计算Neuromorphic computing</font>''')<ref name=":3" /><ref name=":4" /><ref name="humanbrainproject" />是指使用包含电子'''<font color="#ff8000">模拟电路Analog circuit</font>'''的'''<font color="#ff8000">超大规模集成电路Very-large-scale integration</font>'''系统来模拟神经系统中生理结构的研究方法。神经形态计算机或神经形态芯片包括任何使用由硅制成的人造神经元进行计算的设备。<ref name=":5" /><ref name=":2" />近年来，神经形态学（neuromorphic）这个术语被用来描述能够实现'''<font color="#ff8000">神经系统Neural system</font>'''模型功能(如'''<font color="#ff8000">感知Perception</font>'''、'''<font color="#ff8000">运动控制Motor control</font>'''，'''<font color="#ff8000">多感官整合Multisensory integration</font>'''等)的模拟、数字、模拟/数字混合模式超大规模集成电路和软件系统。神经形态计算的硬件实现可以通过基于氧化物的'''<font color="#ff8000">记忆电阻器Memristor</font>'''、<ref name="Maan 1–13" />自旋电子存储器、阈值开关和'''<font color="#ff8000">晶体管Transistor</font>'''来实现。<ref name=":6" /><ref name=":2" />对基于软件的脉冲神经网络系统的训练可以通过误差反向传播机制来实现，例如，使用snnTorch等基于Python的框架，<ref name=":7" />或使用BindsNet等典型的受生物启发的学习模式。<ref name=":8" /><br />
<br />
A key aspect of neuromorphic engineering is understanding how the morphology of individual neurons, circuits, applications, and overall architectures creates desirable computations, affects how information is represented, influences robustness to damage, incorporates learning and development, adapts to local change (plasticity), and facilitates evolutionary change.<br />
<br />
<br />
神经形态工程领域的一个关键问题，就是理解'''<font color="#32CD32">单个神经元形态、神经回路、应用和整体结构</font>'''如何产生理想的计算，如何影响信息的表示和对破坏的鲁棒性，如何整合学习和发展，如何适应局部变化(可塑性) 并促进逐渐发展的变化。<br />
<br />
Neuromorphic engineering is an interdisciplinary subject that takes inspiration from [[biology]], [[physics]], [[mathematics]], [[computer science]], and [[electronic engineering]]<ref name=":2" /> to design artificial neural systems, such as [[Machine vision|vision systems]], head-eye systems, auditory processors, and autonomous robots, whose physical architecture and design principles are based on those of biological nervous systems.<ref name=":9">{{Cite journal | doi = 10.1155/2012/705483| title = Qualitative Functional Decomposition Analysis of Evolved Neuromorphic Flight Controllers| journal = Applied Computational Intelligence and Soft Computing| volume = 2012| pages = 1–21| year = 2012| last1 = Boddhu | first1 = S. K. | last2 = Gallagher | first2 = J. C. | doi-access = free}}</ref> It was developed by [[Carver Mead]]<ref name=":10">{{cite web|title=carver mead website|last1=Mead|first1=Carver|url=http://www.carvermead.caltech.edu/index.html|website=carvermead}}</ref> in the late 1980s.<br />
<br />
<br />
神经形态工程是以'''<font color="#ff8000">生物学Biology</font>'''、'''<font color="#ff8000">物理学Physics</font>'''、'''<font color="#ff8000">数学Mathematics</font>'''、'''<font color="#ff8000">计算机科学Computer science</font>'''和'''<font color="#ff8000">电子工程Electronic engineering<ref name=":2" /></font>'''等学科为基础，设计人工神经系统（如'''<font color="#ff8000">视觉系统Vision systems</font>'''、头眼系统、听觉处理器以及物理结构和设计原则都受启发于生物神经系统的自主机器人）的一门交叉学科。<ref name=":9" />20世纪80年代后期，卡弗·米德极大地推动了神经形态工程领域的发展。<ref name=":10" /><br />
<br />
==Neurological inspiration==<br />
==神经生物学启发==<br />
Neuromorphic engineering is set apart by the inspiration it takes from what we know about the structure and operations of the [[brain]]. Neuromorphic engineering translates what we know about the brain's function into computer systems. Work has mostly focused on replicating the analog nature of [[biological computation]] and the role of [[neuron]]s in [[cognition]].<br />
<br />
<br />
神经形态工程形成的灵感来源于现已知的'''<font color="#ff8000">大脑Brain</font>'''结构及运作机制知识，它将我们对大脑功能的了解用于研究和优化计算机系统。该领域工作主要集中于对'''<font color="#ff8000">生物计算Biological computation</font>'''模拟特性和'''<font color="#ff8000">神经元Neuron</font>'''在'''<font color="#ff8000">认知Cognition</font>'''中所发挥作用的复刻。<br />
<br />
The biological processes of neurons and their [[synapse]]s are dauntingly complex, and thus very difficult to artificially simulate. A key feature of biological brains is that all of the processing in neurons use analog chemical signals. This makes it hard to replicate brains in computers because the current generation of computers is completely digital. However, the characteristics of these parts can be abstracted into mathematical functions that closely capture the essence of the neuron's operations.<br />
<br />
神经元及其'''<font color="#ff8000">突触Synapse</font>'''的生理学过程极其复杂，难以进行人工模拟。神经元在所有处理过程中都使用化学模拟信号，这是大脑的一个关键生理学特征。这个特征大大增加了在计算机中复制大脑的难度，因为目前的计算机是完全数字化的。然而，部分特征可以抽象为数学函数，这些函数能够紧密捕捉神经元运行的本质。<br />
<br />
The goal of neuromorphic computing is not to perfectly mimic the brain and all of its functions, but instead to extract what is known of its structure and operations to be used in a practical computing system. No neuromorphic system will claim nor attempt to reproduce every element of neurons and synapses, but all adhere to the idea that computation is highly [[distributed processing|distributed]] throughout a series of small computing elements analogous to a neuron. While this sentiment is standard, researchers chase this goal with different methods.<ref name=":11">{{Cite journal | doi = 10.1088/1741-2560/13/5/051001| title = Large-scale neuromorphic computing systems| journal = Journal of Neural Engineering| volume = 13| pages = 1–15| year = 2016| last1 = Furber | first1 = Steve| issue = 5| pmid = 27529195| bibcode = 2016JNEng..13e1001F| doi-access = free}}</ref><br />
<br />
神经形态计算的目标不是完美地模拟大脑及其所有功能，而是利用已知的大脑结构和运转机制来研发或优化实际的计算系统。任何神经形态学系统都不会声称或试图复制神经元和突触中的每一个元素，但所有人都一致认可将计算高度'''<font color="#ff8000">分散Distribute</font>'''于一系列类似于神经元的小型计算元素的理念。研究人员用不同的方法来追求这一普遍目标。<ref name=":11" /><br />
<br />
<br />
==Examples==<br />
==代表性成果==<br />
As early as 2006, researchers at [[Georgia Tech]] published a field programmable neural array.<ref name=":12">{{Cite book|title = A field programmable neural array|last1 = Farquhar|first1 = Ethan|date = May 2006|journal = IEEE International Symposium on Circuits and Systems|pages = 4114–4117|last2 = Hasler|first2 = Paul.|doi = 10.1109/ISCAS.2006.1693534|isbn = 978-0-7803-9389-9|s2cid = 206966013}}</ref> This chip was the first in a line of increasingly complex arrays of floating gate transistors that allowed programmability of charge on the gates of [[MOSFET]]s to model the channel-ion characteristics of neurons in the brain and was one of the first cases of a silicon programmable array of neurons.<br />
<br />
早在2006年，佐治亚理工学院的研究人员就研发出了一种现场可编程神经阵列。<ref name=":12" />在此之后，一系列越来越复杂的浮栅晶体管阵列被成功研发出来，这些晶体管阵列可以通过在'''<font color="#ff8000">金属-氧化物半导体效应晶体管MOSFET</font>'''的栅极上编程来模拟大脑中神经元的离子通道特性，这也是以硅为主要材料的可编程神经元阵列的最早成功案例之一。<br />
<br />
In November 2011, a group of [[MIT]] researchers created a computer chip that mimics the analog, ion-based communication in a synapse between two neurons using 400 transistors and standard [[CMOS]] manufacturing techniques.<ref name=":13">{{cite web|title=MIT creates "brain chip"|url=http://www.extremetech.com/extreme/105067-mit-creates-brain-chip|access-date=4 December 2012}}</ref><ref name="Neuromorphic silicon paper">{{cite journal|title=Neuromorphic silicon neurons and large-scale neural networks: challenges and opportunities|doi=10.3389/fnins.2011.00108|pmid=21991244|pmc=3181466|volume=5|pages=108|journal=Frontiers in Neuroscience|year=2011|last1=Poon|first1=Chi-Sang|last2=Zhou|first2=Kuan|doi-access=free}}</ref><br />
<br />
2011年11月，麻省理工学院的一组研究人员研发出一种计算机芯片，该芯片上使用标准的'''<font color="#ff8000">互补金属氧化物半导体CMOS</font>'''制造技术集成了400个晶体管，能够模拟神经元间突触中基于离子的通讯。<ref name=":13" /><ref name="Neuromorphic silicon paper" /><br />
<br />
In June 2012, [[spintronic]] researchers at [[Purdue University]] presented a paper on the design of a neuromorphic chip using [[Spin valve|lateral spin valve]]s and [[memristor]]s. They argue that the architecture works similarly to neurons and can therefore be used to test methods of reproducing the brain's processing. In addition, these chips are significantly more energy-efficient than conventional ones.<ref name="Spin Devices Prop">{{Cite arXiv|title=Proposal For Neuromorphic Hardware Using Spin Devices|eprint=1206.3227|last1=Sharad|first1=Mrigank|last2=Augustine|first2=Charles|last3=Panagopoulos|first3=Georgios|last4=Roy|first4=Kaushik|class=cond-mat.dis-nn|year=2012}}</ref><br />
<br />
2012年6月，普渡大学的'''<font color="#ff8000">自旋电子学Spintronic</font>'''研究人员发表了一篇关于利用'''<font color="#ff8000">侧向自旋阀Lateral spin valves</font>'''和'''<font color="#ff8000">记忆电阻器Memristors</font>'''设计神经形态芯片的论文。他们认为，这种芯片结构的工作原理与神经元相似，因此可以用于大脑运行机制的复刻方法的测试。此外，这些芯片在能耗方面明显优于传统芯片。<ref name="Spin Devices Prop" /><br />
<br />
Research at [[HP Labs]] on Mott memristors has shown that while they can be non-[[Volatile memory|volatile]], the volatile behavior exhibited at temperatures significantly below the [[phase transition]] temperature can be exploited to fabricate a [[neuristor]],<ref name=":0" /> a biologically-inspired device that mimics behavior found in neurons.<ref name=":0">{{Cite journal | doi = 10.1038/nmat3510| pmid = 23241533| title = A scalable neuristor built with Mott memristors| journal = Nature Materials| volume = 12| issue = 2| pages = 114–7| year = 2012| last1 = Pickett | first1 = M. D. | last2 = Medeiros-Ribeiro | first2 = G. | last3 = Williams | first3 = R. S. | bibcode = 2013NatMa..12..114P| s2cid = 16271627| url = https://semanticscholar.org/paper/b6ba6f496ace2b947f111059663e76bb60e9efeb}}</ref> In September 2013, they presented models and simulations that show how the spiking behavior of these neuristors can be used to form the components required for a [[Turing machine]].<ref name=":14">{{cite journal|doi=10.1088/0957-4484/24/38/384002|title=Phase transitions enable computational universality in neuristor-based cellular automata|author1=Matthew D Pickett|author2=R Stanley Williams|name-list-style=amp|date=September 2013|publisher=IOP Publishing Ltd|journal=Nanotechnology|volume=24|issue=38|pmid=23999059|bibcode=2013Nanot..24L4002P|at=384002}}</ref><br />
<br />
'''<font color="#ff8000">惠普实验室HP labs</font>'''在莫特记忆电阻器上的研究表明，尽管它们可以是非'''<font color="#ff8000">易失性Volatile</font>'''的，但是在'''<font color="#ff8000">相变Phase transition</font>'''温度以下时表现出的易失性行为可以被用来制造'''<font color="#ff8000">类神经元电阻器Neuristor</font>'''(一种生物学启发的模仿神经元行为的硬件)<ref name=":0" />。2013年9月，他们通过模型和仿真展示了这些类神经元电阻器的脉冲行为如何产生'''<font color="#ff8000">图灵机Turing machine</font>'''的所需元素。<ref name=":14" /><br />
<br />
[[Neurogrid]], built by ''Brains in Silicon'' at [[Stanford University]],<ref name=":15">{{cite journal|last1=Boahen|first1=Kwabena|title=Neurogrid: A Mixed-Analog-Digital Multichip System for Large-Scale Neural Simulations|journal=Proceedings of the IEEE|date=24 April 2014|volume=102|issue=5|pages=699–716|doi=10.1109/JPROC.2014.2313565|s2cid=17176371}}</ref> is an example of hardware designed using neuromorphic engineering principles. The circuit board is composed of 16 custom-designed chips, referred to as NeuroCores. Each NeuroCore's analog circuitry is designed to emulate neural elements for 65536 neurons, maximizing energy efficiency. The emulated neurons are connected using digital circuitry designed to maximize spiking throughput.<ref name=":16">{{cite journal|doi=10.1038/503022a|pmid = 24201264|title = Neuroelectronics: Smart connections|journal = Nature|volume = 503|issue = 7474|pages = 22–4|year = 2013|last1 = Waldrop|first1 = M. Mitchell|bibcode = 2013Natur.503...22W|doi-access = free}}</ref><ref name=":17">{{cite journal|doi=10.1109/JPROC.2014.2313565|title = Neurogrid: A Mixed-Analog-Digital Multichip System for Large-Scale Neural Simulations|journal = Proceedings of the IEEE|volume = 102|issue = 5|pages = 699–716|year = 2014|last1 = Benjamin|first1 = Ben Varkey|last2 = Peiran Gao|last3 = McQuinn|first3 = Emmett|last4 = Choudhary|first4 = Swadesh|last5 = Chandrasekaran|first5 = Anand R.|last6 = Bussat|first6 = Jean-Marie|last7 = Alvarez-Icaza|first7 = Rodrigo|last8 = Arthur|first8 = John V.|last9 = Merolla|first9 = Paul A.|last10 = Boahen|first10 = Kwabena|s2cid = 17176371}}</ref><br />
<br />
'''<font color="#ff8000">神经栅格Neurogrid<ref name=":15" /></font>'''是由斯坦福大学Brains in Silicon公司研发的、使用神经形态工程原理设计的硬件。该电路板由16个定制设计的芯片组成(NeuroCores)。在设计中，每个NeuroCore芯片的模拟电路对65536个神经元的神经元素进行模拟，以最大限度地提高能量效率。模拟出的神经元通过设计的数字电路连接，以最大化脉冲吞吐量。<ref name=":16" /><ref name=":17" /><br />
<br />
A research project with implications for neuromorphic engineering is the [[Human Brain Project]] that is attempting to simulate a complete human brain in a supercomputer using biological data. It is made up of a group of researchers in neuroscience, medicine, and computing.<ref name=":18">{{cite web|title=Involved Organizations|url=http://www.humanbrainproject.eu/partners.html|access-date=22 February 2013|url-status=dead|archive-url=https://web.archive.org/web/20130302142627/http://www.humanbrainproject.eu/partners.html|archive-date=2 March 2013}}</ref> [[Henry Markram]], the project's co-director, has stated that the project proposes to establish a foundation to explore and understand the brain and its diseases, and to use that knowledge to build new computing technologies. The three primary goals of the project are to better understand how the pieces of the brain fit and work together, to understand how to objectively diagnose and treat brain diseases, and to use the understanding of the human brain to develop neuromorphic computers. That the simulation of a complete human brain will require a supercomputer a thousand times more powerful than today's encourages the current focus on neuromorphic computers.<ref name=":19">{{cite web|title=Human Brain Project|url=http://www.humanbrainproject.eu|access-date=22 February 2013}}</ref> $1.3 billion has been allocated to the project by The [[European Commission]].<ref name=":20">{{cite web|title=The Human Brain Project and Recruiting More Cyberwarriors|url=http://www.marketplace.org/topics/tech/human-brain-project-and-recruiting-more-cyberwarriors|access-date=22 February 2013|date=January 29, 2013}}</ref><br />
<br />
'''<font color="#ff8000">人类大脑计划Human Brain Project</font>'''对神经形态工程具有较大影响，其主要任务是尝试用生物数据在超级计算机中模拟完整的人脑。人类大脑计划由神经科学、医学和计算机科学背景的研究人员组成。<ref name=":18" />该项目的联合主管亨利•马克拉姆(Henry Markram)表示，人类大脑计划的目的是建立一个探索和了解脑科学和脑疾病知识的基础，并利用这些知识来构建更先进的计算机技术。这个项目的三个主要目标分别是: 更好地理解大脑的各个部分是如何相互配合协同工作的; 理解如何客观地诊断和治疗脑部疾病; 以及利用对人类大脑的理解来开发神经形态计算机。模拟一个完整的人类大脑需要一台比现在强大一千倍的超级计算机，这不断激发着对神经形态计算机领域的研究兴趣。<ref name=":19" />欧盟委员会已经向人类大脑计划拨款13亿美元。<ref name=":20" /><br />
<br />
Other research with implications for neuromorphic engineering involves the [[BRAIN Initiative]]<ref name="economist">[https://www.economist.com/news/science-and-technology/21582495-computers-will-help-people-understand-brains-better-and-understanding-brains Neuromorphic computing: The machine of a new soul], The Economist, 2013-08-03</ref> and the [[TrueNorth]] chip from [[IBM]].<ref name=":21">{{cite journal|last1=Modha|first1=Dharmendra|title=A million spiking-neuron integrated circuit with a scalable communication network and interface|journal=Science|date=Aug 2014|volume=345|issue=6197|pages=668–673|doi=10.1126/science.1254642|pmid=25104385|bibcode=2014Sci...345..668M|s2cid=12706847}}</ref> Neuromorphic devices have also been demonstrated using nanocrystals, nanowires, and conducting polymers.<ref name=":22">{{Cite web|url=http://jessamynfairfield.com/wp-content/uploads/2017/03/PWMar17Fairfield.pdf|title=Smarter Machines|last=Fairfield|first=Jessamyn|date=March 1, 2017}}</ref><br />
<br />
其他与神经形态工程有关的研究还包括'''<font color="#ff8000">脑计划BRAIN initiative</font>'''，<ref name="economist" />和IBM研发的'''<font color="#ff8000">TrueNorth</font>'''芯片。<ref name=":21" />使用纳米晶体、纳米线和导电聚合物也能够用于制造神经形态学硬件。<ref name=":22" /><br />
<br />
[[Intel]] unveiled its neuromorphic research chip, called “[[Intel Loihi|Loihi]]”, in October 2017. The chip uses an asynchronous [[spiking neural network]] (SNN) to implement adaptive self-modifying event-driven fine-grained parallel computations used to implement learning and inference with high efficiency.<ref name=":23">{{cite journal |last1=Davies |first1=Mike |title=Loihi: A Neuromorphic Manycore Processor with On-Chip Learning |journal=IEEE Micro |date=January 16, 2018 |volume=38 |issue=1 |pages=82–99 |display-authors=etal|doi=10.1109/MM.2018.112130359 |s2cid=3608458 }}</ref><ref name="Morris2017">{{cite web |last1=Morris |first1=John |title=Why Intel built a neuromorphic chip |url=https://www.zdnet.com/article/why-intel-built-a-neuromorphic-chip/ |website=ZDNet |access-date=17 August 2018 |language=en}}</ref><br />
<br />
2017年10月，英特尔发布了神经形态芯片'''<font color="#ff8000">Loihi</font>'''。该芯片采用异步'''<font color="#ff8000">脉冲神经网络Spiking neural network</font>'''实现了自适应、自修改、事件驱动的细粒度并行计算，实现了高效的学习和推理。<ref name=":23" /><ref name="Morris2017" /><br />
<br />
[[IMEC]], a Belgium-based nanoelectronics research center, demonstrated the world's first self-learning neuromorphic chip. The brain-inspired chip, based on OxRAM technology, has the capability of self-learning and has been demonstrated to have the ability to compose music.<ref name=":24">{{cite web |title=Imec demonstrates self-learning neuromorphic chip that composes music |url=https://www.imec-int.com/en/articles/imec-demonstrates-self-learning-neuromorphic-chip-that-composes-music |website=IMEC International |access-date=1 October 2019}}</ref> IMEC released the 30-second tune composed by the prototype. The chip was sequentially loaded with songs in the same time signature and style. The songs were old Belgian and French flute minuets, from which the chip learned the rules at play and then applied them.<ref name=":25">{{cite web|last1=Bourzac|first1=Katherine|title=A Neuromorphic Chip That Makes Music|url=https://spectrum.ieee.org/a-neuromorphic-chip-that-makes-music|url-status=live|access-date=1 October 2019|website=IEEE Spectrum|date=May 23, 2017}}</ref><br />
<br />
比利时的'''<font color="#ff8000">微电子研究中心IMEC</font>'''研发了世界上首个自学习神经形态芯片。这种基于'''<font color="#32CD32"> OxRAM(filamentary-oxide-based resistive memory technology)</font>''' 技术的大脑启发芯片具有自学习能力，并且已被证明具有创作音乐的能力。<ref name=":24" />IMEC发布了由芯片原型机谱写的30秒旋律。向芯片加载一系列特征、风格统一的歌曲（古代比利时和法国长笛小步舞曲），芯片从中学习相关规则并将其应用于创作。<ref name=":25" /><br />
<br />
[[Blue Brain Project|The Blue Brain Project]], led by Henry Markram, aims to build biologically detailed digital reconstructions and simulations of the mouse brain. The Blue Brain Project has created in silico models of rodent brains, while attempting to replicate as many details about its biology as possible. The supercomputer-based simulations offer new perspectives on understanding the structure and functions of the brain.<br />
<br />
由亨利·马克拉姆领导的'''<font color="#ff8000">蓝脑计划The Blue Brain Project</font>'''旨在建立小鼠大脑生理学细节的数字重建和模拟。蓝脑计划已经建立了啮齿动物大脑的电子模型，同时进行着尽可能多地复制其生理学细节的尝试。基于超级计算机的模拟为理解大脑的结构和功能提供了新的视角。<br />
<br />
The European Union funded a series of projects at the University of Heidelberg, which led to the development of <br />
[[BrainScaleS]] (brain-inspired multiscale computation in neuromorphic hybrid systems), a hybrid analog [[neuromorphic]] supercomputer located at Heidelberg University, Germany. It was developed as part of the [[Human Brain Project]] neuromorphic computing platform and is the complement to the [[SpiNNaker]] supercomputer (which is based on digital technology). The architecture used in BrainScaleS mimics biological neurons and their connections on a physical level; additionally, since the components are made of silicon, these model neurons operate on average 864 times (24 hours of real time is 100 seconds in the machine simulation) that of their biological counterparts.<ref name=":26">{{Cite web|date=2016-03-21|title=Beyond von Neumann, Neuromorphic Computing Steadily Advances|url=https://www.hpcwire.com/2016/03/21/lacking-breakthrough-neuromorphic-computing-steadily-advance/|access-date=2021-10-08|website=HPCwire|language=en-US}}</ref><br />
<br />
欧盟资助了海德堡大学一系列促进BrainScaleS(神经形态混合系统中受大脑启发的多尺度计算)发展的项目，这是一台位于德国海德堡大学的混合模拟'''<font color="#ff8000">神经形态Neuromorphic</font>'''超级计算机。它是作为人类大脑计划中神经形态计算平台的一部分而开发的，是'''<font color="#ff8000">SpiNNaker</font>'''超级计算机(基于数字技术)的补充。BrainScaleS中使用的体系架构模拟了生物神经元及其在物理层面上的连接；此外，由于这些组件是由硅制成的，这些模型神经元平均运行速度是生物神经元的864倍，这意味着在机器模拟中，24小时的实时时间仅为100秒。<ref name=":26" /><br />
<br />
===Neuromorphic sensors===<br />
===神经形态传感器===<br />
The concept of neuromorphic systems can be extended to sensors (not just to computation). An example of this applied to detecting [[light]] is the [[retinomorphic sensor]] or, when employed in an array, the [[event camera]].<br />
<br />
神经形态系统的概念可以扩展到传感器(而不仅仅是计算单元)。一个用于检测光线的例子是'''<font color="#ff8000">类视网膜传感器Retinomorphic sensor</font>'''，或者'''<font color="#ff8000">事件摄像机Event camera</font>阵列。'''<br />
<br />
<br />
==Ethical considerations ==<br />
==伦理问题==<br />
While the interdisciplinary concept of neuromorphic engineering is relatively new, many of the same ethical considerations apply to neuromorphic systems as apply to [[human-like machines]] and [[artificial intelligence]] in general. However, the fact that neuromorphic systems are designed to mimic a [[human brain]] gives rise to unique ethical questions surrounding their usage.<br />
<br />
虽然神经形态工程这个跨学科概念相对较新，但许多适用于'''<font color="#ff8000">类人机器Human-like machines</font>'''和人工智能的伦理考虑大体上也适用于神经形态系统。另一方面，神经形态系统是为了模仿人类大脑而设计，这一事实引起了一些针对性的独特伦理问题。<br />
<br />
However, the practical debate is that neuromorphic hardware as well as artificial "neural networks" are immensely simplified models of how the brain operates or processes information at a much lower [[complex system|complexity in terms of size and functional technology]] and a much more regular structure in terms of [[brain connectivity|connectivity]]. Comparing [[neuromorphic chip]]s to the brain is a very crude comparison similar to comparing a plane to a bird just because they both have wings and a tail. The fact is that neural cognitive systems are many orders of magnitude more [[Energy efficiency (physics)|energy-]] and compute-efficient than current state-of-the-art AI and neuromorphic engineering is an attempt to narrow this gap by inspiring from the brain's mechanism just like many engineering designs have [[bioengineering|bio-inspired features]].<br />
<br />
然而，这种争论的实际情况是，神经形态硬件和人工“神经网络”是对大脑运作或处理信息过程的极其简化的模型，在大小和功能技术方面的'''<font color="#ff8000">复杂性Complexity</font>'''相比而言比较低，在'''<font color="#ff8000">连接性Connectivity</font>'''方面也具有更加规则的结构。将'''<font color="#ff8000">神经形态芯片Neuromorphic chips</font>'''与大脑进行比较是一种非常粗糙的比较，类似于仅仅因为飞机有翅膀和尾巴就将它与鸟进行比较。事实上，当前最先进的人工智能在'''<font color="#ff8000">能耗效率Energy-efficiency</font>'''和计算效率方面距离人脑神经认知系统仍有较大差距，而神经形态工程只是一种通过从大脑机制中获得灵感来缩小这种差距的尝试，就像许多工程设计中都具有'''<font color="#ff8000">生物启发的特征Bio-inspired features</font>'''一样。<br />
<br />
===Democratic concerns===<br />
===公众担忧===<br />
Significant ethical limitations may be placed on neuromorphic engineering due to public perception.<ref>{{Cite report|url=https://ai100.stanford.edu/sites/g/files/sbiybj9861/f/ai_100_report_0831fnl.pdf|title=Artificial Intelligence and Life in 2030|author=2015 Study Panel|date=September 2016|work=One Hundred Year Study on Artificial Intelligence (AI100)|publisher=Stanford University}}</ref> Special [[Eurobarometer]] 382: Public Attitudes Towards Robots, a survey conducted by the European Commission, found that 60% of [[European Union]] citizens wanted a ban of robots in the care of children, the elderly, or the disabled. Furthermore, 34% were in favor of a ban on robots in education, 27% in healthcare, and 20% in leisure. The European Commission classifies these areas as notably “human.” The report cites increased public concern with robots that are able to mimic or replicate human functions. Neuromorphic engineering, by definition, is designed to replicate the function of the human brain.<ref name=":1">{{Cite web|url=http://ec.europa.eu/commfrontoffice/publicopinion/archives/ebs/ebs_382_en.pdf|title=Special Eurobarometer 382: Public Attitudes Towards Robots|last=European Commission|date=September 2012|website=European Commission}}</ref><br />
<br />
由于公众认知的相关忧虑，神经形态工程学可能会受到严重的伦理限制。欧盟委员会进行的一项调查发现，60% 的欧盟公民希望禁止机器人参与照顾儿童、老人或残疾人的工作。此外，34% 的人支持禁止机器人用于教育，27% 的人支持禁止机器人用于医疗保健，20% 的人支持禁止机器人用于休闲。欧盟委员会将以上领域划入“人类”范畴。报告指出，公众越来越关注能够模仿或复制人类功能的机器人。神经形态工程，顾名思义，是为了复制人脑的功能而设计的。<br />
<br />
The democratic concerns surrounding neuromorphic engineering are likely to become even more profound in the future. The European Commission found that EU citizens between the ages of 15 and 24 are more likely to think of robots as human-like (as opposed to instrument-like) than EU citizens over the age of 55. When presented an image of a robot that had been defined as human-like, 75% of EU citizens aged 15–24 said it corresponded with the idea they had of robots while only 57% of EU citizens over the age of 55 responded the same way. The human-like nature of neuromorphic systems, therefore, could place them in the categories of robots many EU citizens would like to see banned in the future.<ref name=":1" /><br />
<br />
The democratic concerns surrounding neuromorphic engineering are likely to become even more profound in the future. The European Commission found that EU citizens between the ages of 15 and 24 are more likely to think of robots as human-like (as opposed to instrument-like) than EU citizens over the age of 55. When presented an image of a robot that had been defined as human-like, 75% of EU citizens aged 15–24 said it corresponded with the idea they had of robots while only 57% of EU citizens over the age of 55 responded the same way. The human-like nature of neuromorphic systems, therefore, could place them in the categories of robots many EU citizens would like to see banned in the future.<br />
<br />
围绕神经形态工程的公众担忧可能在未来变得更加深刻。欧盟委员会发现，15至24岁的欧盟公民比55岁以上的欧盟公民更有可能认为机器人像人(而不是像仪器)。当看到一张被定义为“类人”机器人的图片时，年龄在15岁至24岁之间的欧盟公民中有75% 的人表示，这与他们对机器人的想法相符，而55岁以上的欧盟公民中只有57% 的人有同样的反应。因此，类似人类的神经形态系统，可以把它们归入许多欧盟公民希望在未来禁止使用的机器人类别。<br />
<br />
===Personhood===<br />
=== 人格问题===<br />
As neuromorphic systems have become increasingly advanced, some scholars{{who|date=August 2021}} have advocated for granting [[personhood]] rights to these systems. If the brain is what grants humans their personhood, to what extent does a neuromorphic system have to mimic the human brain to be granted personhood rights? Critics of technology development in the [[Human Brain Project]], which aims to advance brain-inspired computing, have argued that advancement in neuromorphic computing could lead to machine consciousness or personhood.<ref>{{Cite journal|last=Aicardi|first=Christine|date=September 2018|title=Accompanying technology development in the Human Brain Project: From foresight to ethics management|journal=Futures|volume=102|pages=114–124|doi=10.1016/j.futures.2018.01.005|doi-access=free}}</ref> If these systems are to be treated as people, critics argue, then many tasks humans perform using neuromorphic systems, including the act of termination of neuromorphic systems, may be morally impermissible as these acts would violate the autonomy of the neuromorphic systems.<ref>{{Cite journal|last=Lim|first=Daniel|date=2014-06-01|title=Brain simulation and personhood: a concern with the Human Brain Project|journal=Ethics and Information Technology|language=en|volume=16|issue=2|pages=77–89|doi=10.1007/s10676-013-9330-5|s2cid=17415814|issn=1572-8439}}</ref><br />
<br />
As neuromorphic systems have become increasingly advanced, some scholars have advocated for granting personhood rights to these systems. If the brain is what grants humans their personhood, to what extent does a neuromorphic system have to mimic the human brain to be granted personhood rights? Critics of technology development in the Human Brain Project, which aims to advance brain-inspired computing, have argued that advancement in neuromorphic computing could lead to machine consciousness or personhood. If these systems are to be treated as people, critics argue, then many tasks humans perform using neuromorphic systems, including the act of termination of neuromorphic systems, may be morally impermissible as these acts would violate the autonomy of the neuromorphic systems.<br />
<br />
随着神经形态系统的日益发展，一些学者主张赋予这些系统人格权。如果是大脑赋予了人类人格，那么神经形态系统在多大程度上必须模仿人类大脑才能被赋予人格权利？“人脑计划”旨在推进以大脑为灵感的计算机技术发展，该计划的批评者认为，神经形态计算机技术的进步可能导致机器意识或人格的形成。批评家认为，如果这些系统被当作人来对待，那么人类使用神经形态系统执行的许多任务，包括终止神经形态系统的行为，可能在道德上是不允许的，因为这些行为将违反神经形态系统的自主性。<br />
<br />
<br />
==Dual use (military applications) ==<br />
==军民两用技术==<br />
The [[Joint Artificial Intelligence Center]], a branch of the U.S. military, is a center dedicated to the procurement and implementation of AI software and neuromorphic hardware for combat use. Specific applications include smart headsets/goggles and robots. JAIC intends to rely heavily on neuromorphic technology to connect "every fighter every shooter" within a network of neuromorphic-enabled units.<br />
<br />
The Joint Artificial Intelligence Center, a branch of the U.S. military, is a center dedicated to the procurement and implementation of AI software and neuromorphic hardware for combat use. Specific applications include smart headsets/goggles and robots. JAIC intends to rely heavily on neuromorphic technology to connect "every fighter every shooter" within a network of neuromorphic-enabled units.<br />
<br />
军民两用联合人工智能中心是美国军队的一个分支，专门从事采购和实施用于战斗的人工智能软件和神经形态硬件。具体应用包括智能耳机/护目镜和机器人。JAIC 打算严重依赖神经形态技术来连接神经形态单位网络中的“每个战士每个射手”。<br />
<br />
==Legal considerations==<br />
==法律问题==<br />
Skeptics have argued that there is no way to apply the electronic personhood, the concept of personhood that would apply to neuromorphic technology, legally. In a letter signed by 285 experts in law, robotics, medicine, and ethics opposing a European Commission proposal to recognize “smart robots” as legal persons, the authors write, “A legal status for a robot can’t derive from the [[Natural person|Natural Person]] model, since the robot would then hold [[human rights]], such as the right to dignity, the right to its integrity, the right to remuneration or the right to citizenship, thus directly confronting the Human rights. This would be in contradiction with the [[Charter of Fundamental Rights of the European Union]] and the [[Convention for the Protection of Human Rights and Fundamental Freedoms]].”<ref>{{Cite web|url=http://www.robotics-openletter.eu/|title=Robotics Openletter {{!}} Open letter to the European Commission|language=fr-FR|access-date=2019-05-10}}</ref><br />
<br />
Skeptics have argued that there is no way to apply the electronic personhood, the concept of personhood that would apply to neuromorphic technology, legally. In a letter signed by 285 experts in law, robotics, medicine, and ethics opposing a European Commission proposal to recognize “smart robots” as legal persons, the authors write, “A legal status for a robot can’t derive from the Natural Person model, since the robot would then hold human rights, such as the right to dignity, the right to its integrity, the right to remuneration or the right to citizenship, thus directly confronting the Human rights. This would be in contradiction with the Charter of Fundamental Rights of the European Union and the Convention for the Protection of Human Rights and Fundamental Freedoms.”<br />
<br />
怀疑论者认为，在法律上没有办法应用电子人格，这个人格概念将适用于神经形态技术。在一封由285名法律、机器人、医学和伦理学专家签名的信中，作者们反对欧盟委员会承认“智能机器人”为法人的提议，他们写道，“机器人的法律地位不能从自然人模型中推导出来，因为机器人将拥有人权，如尊严权、完整权、报酬权或公民权，从而直接面对人权。这将有悖于《欧洲联盟基本权利宪章和《保护人权和基本自由公约》。”<br />
<br />
===Ownership and property rights ===<br />
===所有权及财产权问题===<br />
There is significant legal debate around property rights and artificial intelligence. In ''Acohs Pty Ltd v. Ucorp Pty Ltd'', Justice Christopher Jessup of the [[Federal Court of Australia]] found that the [[source code]] for [[Material safety data sheets|Material Safety Data Sheets]] could not be [[Copyright law of Australia|copyrighted]] as it was generated by a [[software interface]] rather than a human author.<ref>{{Cite web|url=http://www.lavan.com.au/advice/intellectual_property_technology/copyright_in_source_code_and_digital_products|title=Copyright in source code and digital products|last=Lavan|website=Lavan|language=en|access-date=2019-05-10}}</ref> The same question may apply to neuromorphic systems: if a neuromorphic system successfully mimics a human brain and produces a piece of original work, who, if anyone, should be able to claim ownership of the work?<ref>{{cite journal |last1=Eshraghian|first1=Jason K. |title=Human Ownership of Artificial Creativity |journal=Nature Machine Intelligence |date=9 March 2020 |volume=2 |pages=157–160 |doi=10.1038/s42256-020-0161-x}}</ref><br />
<br />
There is significant legal debate around property rights and artificial intelligence. In Acohs Pty Ltd v. Ucorp Pty Ltd, Justice Christopher Jessup of the Federal Court of Australia found that the source code for Material Safety Data Sheets could not be copyrighted as it was generated by a software interface rather than a human author. The same question may apply to neuromorphic systems: if a neuromorphic system successfully mimics a human brain and produces a piece of original work, who, if anyone, should be able to claim ownership of the work?<br />
<br />
围绕财产权和人工智能有着重大的法律争论。在 Acohs Pty Ltd 诉 Ucorp Pty Ltd 一案中，澳大利亚联邦法院的克里斯托弗 · 杰瑟普法官发现，材料安全数据表的源代码不能受版权保护，因为它是由软件界面而不是人工作者生成的。同样的问题可能也适用于神经形态系统: 如果一个神经形态系统成功地模仿了人类的大脑并产生了一部原创作品，那么谁，如果有人，应该声称拥有这部作品的所有权？<br />
<br />
==Neuromemristive systems==<br />
== 神经记忆电阻系统==<br />
Neuromemristive systems are a subclass of neuromorphic computing systems that focus on the use of [[memristors]] to implement [[neuroplasticity]]. While neuromorphic engineering focuses on mimicking biological behavior, neuromemristive systems focus on abstraction.<ref>{{Cite web|url=https://digitalops.sandia.gov/Mediasite/Play/a10cf6ceb55d47608bb8326dd00e46611d|title=002.08 N.I.C.E. Workshop 2014: Towards Intelligent Computing with Neuromemristive Circuits and Systems - Feb. 2014|website=digitalops.sandia.gov|access-date=2019-08-26}}</ref> For example, a neuromemristive system may replace the details of a [[Cerebral cortex|cortical]] microcircuit's behavior with an abstract neural network model.<ref>C. Merkel and D. Kudithipudi, "Neuromemristive extreme learning machines for pattern classification," ISVLSI, 2014.</ref><br />
<br />
Neuromemristive systems are a subclass of neuromorphic computing systems that focus on the use of memristors to implement neuroplasticity. While neuromorphic engineering focuses on mimicking biological behavior, neuromemristive systems focus on abstraction. For example, a neuromemristive system may replace the details of a cortical microcircuit's behavior with an abstract neural network model.C. Merkel and D. Kudithipudi, "Neuromemristive extreme learning machines for pattern classification," ISVLSI, 2014.<br />
<br />
神经记忆电阻系统是神经形态计算系统的一个亚类，主要研究利用记忆电阻器实现神经可塑性。神经形态工程的重点是模拟生物行为，而神经记忆电阻系统的重点是提取。例如，一个神经记忆系统可以用一个抽象的神经网络模型取代皮层微电路的行为细节。默克尔和 d. Kudithipudi，“用于模式分类的神经记忆极端学习机器”，ISVLSI，2014。<br />
<br />
There exist several neuron inspired threshold logic functions<ref name="Maan 1–13" /> implemented with memristors that have applications in high level [[pattern recognition]] applications. Some of the applications reported recently include [[speech recognition]],<ref>{{Cite journal|title = Memristor pattern recogniser: isolated speech word recognition|journal = Electronics Letters|pages = 1370–1372|volume = 51|issue = 17|doi = 10.1049/el.2015.1428|first1 = A.K.|last1 = Maan|first2 = A.P.|last2 = James|first3 = S.|last3 = Dimitrijev|year = 2015|bibcode = 2015ElL....51.1370M|hdl = 10072/140989|s2cid = 61454815|url = https://semanticscholar.org/paper/48d3ab11ec6e213b62f11eedcfb7b7febb058674|hdl-access = free}}</ref> [[face recognition]]<ref>{{Cite journal|title = Memristive Threshold Logic Face Recognition|journal = Procedia Computer Science|date = 2014-01-01|pages = 98–103|volume = 41|series = 5th Annual International Conference on Biologically Inspired Cognitive Architectures, 2014 BICA|doi = 10.1016/j.procs.2014.11.090|first1 = Akshay Kumar|last1 = Maan|first2 = Dinesh S.|last2 = Kumar|first3 = Alex Pappachen|last3 = James|doi-access = free}}</ref> and [[object recognition]].<ref>{{Cite journal|title = Memristive Threshold Logic Circuit Design of Fast Moving Object Detection|journal = IEEE Transactions on Very Large Scale Integration (VLSI) Systems|date = 2015-10-01|issn = 1063-8210|pages = 2337–2341|volume = 23|issue = 10|doi = 10.1109/TVLSI.2014.2359801|first1 = A.K.|last1 = Maan|first2 = D.S.|last2 = Kumar|first3 = S.|last3 = Sugathan|first4 = A.P.|last4 = James|arxiv = 1410.1267|s2cid = 9647290}}</ref> They also find applications in replacing conventional digital logic gates.<ref>{{Cite journal|title = Resistive Threshold Logic|journal = IEEE Transactions on Very Large Scale Integration (VLSI) Systems|date = 2014-01-01|issn = 1063-8210|pages = 190–195|volume = 22|issue = 1|doi = 10.1109/TVLSI.2012.2232946|first1 = A.P.|last1 = James|first2 = L.R.V.J.|last2 = Francis|first3 = D.S.|last3 = Kumar|arxiv = 1308.0090|s2cid = 7357110}}</ref><ref>{{Cite journal|title = Threshold Logic Computing: Memristive-CMOS Circuits for Fast Fourier Transform and Vedic Multiplication|journal = IEEE Transactions on Very Large Scale Integration (VLSI) Systems|date = 2015-11-01|issn = 1063-8210|pages = 2690–2694|volume = 23|issue = 11|doi = 10.1109/TVLSI.2014.2371857|first1 = A.P.|last1 = James|first2 = D.S.|last2 = Kumar|first3 = A.|last3 = Ajayan|arxiv = 1411.5255|s2cid = 6076956}}</ref><br />
<br />
There exist several neuron inspired threshold logic functions implemented with memristors that have applications in high level pattern recognition applications. Some of the applications reported recently include speech recognition, face recognition and object recognition. They also find applications in replacing conventional digital logic gates.<br />
<br />
利用记忆电阻器实现的神经元启发阈值逻辑函数在高级模式识别中有着广泛的应用。最近报道的一些应用包括语音识别、人脸识别和物体识别。它们还可以用来取代传统的数字逻辑门。<br />
<br />
For ideal passive memristive circuits there is an exact equation (Caravelli-Traversa-[[Di Ventra]] equation) for the internal memory of the circuit:<ref>{{cite journal |last=Caravelli |display-authors=etal|arxiv=1608.08651 |title=The complex dynamics of memristive circuits: analytical results and universal slow relaxation |year=2017 |doi=10.1103/PhysRevE.95.022140 |pmid= 28297937 |volume=95 |issue= 2 |pages= 022140 |journal=Physical Review E|bibcode=2017PhRvE..95b2140C |s2cid=6758362}}</ref><br />
<br />
For ideal passive memristive circuits there is an exact equation (Caravelli-Traversa-Di Ventra equation) for the internal memory of the circuit:<br />
<br />
对于理想的无源记忆电路，电路的内部记忆有一个精确的方程(Caravelli-Traversa-Di Ventra 方程) :<br />
<br />
:<math> \frac{d}{dt} \vec{W} = \alpha \vec{W}-\frac{1}{\beta} (I+\xi \Omega W)^{-1} \Omega \vec S </math><br />
<br />
:\frac{d}{dt} \vec{W} = \alpha \vec{W}-\frac{1}{\beta} (I+\xi \Omega W)^{-1} \Omega \vec S<br />
<br />
:\frac{d}{dt} \vec{W} = \alpha \vec{W}-\frac{1}{\beta} (I+\xi \Omega W)^{-1} \Omega \vec S<br />
<br />
as a function of the properties of the physical memristive network and the external sources. In the equation above, <math>\alpha</math> is the "forgetting" time scale constant, <math> \xi=r-1</math> and <math>r=\frac{R_\text{off}}{R_\text{on}}</math> is the ratio of ''off'' and ''on'' values of the limit resistances of the memristors, <math> \vec S </math> is the vector of the sources of the circuit and <math>\Omega</math> is a projector on the fundamental loops of the circuit. The constant <math>\beta</math> has the dimension of a voltage and is associated to the properties of the [[memristor]]; its physical origin is the charge mobility in the conductor. The diagonal matrix and vector <math>W=\operatorname{diag}(\vec W)</math> and <math>\vec W</math> respectively, are instead the internal value of the memristors, with values between 0 and 1. This equation thus requires adding extra constraints on the memory values in order to be reliable.<br />
<br />
<nowiki>as a function of the properties of the physical memristive network and the external sources. In the equation above, \alpha is the "forgetting" time scale constant, \xi=r-1 and r=\frac{R_\text{off}}{R_\text{on}} is the ratio of off and on values of the limit resistances of the memristors, \vec S is the vector of the sources of the circuit and \Omega is a projector on the fundamental loops of the circuit. The constant \beta has the dimension of a voltage and is associated to the properties of the memristor; its physical origin is the charge mobility in the conductor. The diagonal matrix and vector W=\operatorname{diag}(\vec W) and \vec W respectively, are instead the internal value of the memristors, with values between 0 and 1. This equation thus requires adding extra constraints on the memory values in order to be reliable.</nowiki><br />
<br />
<nowiki>作为物理记忆网络和外部源的性质的函数。在上述方程中，α 是“遗忘”时间尺度常数，xi = r-1，r = frac { r _ text { off }{ on }{ r _ text { on }}是记忆电阻器极限电阻的开关和开关值之比，vec s 是电路源的矢量，Omega 是电路基本环路的投影仪。常数 β 具有电压的尺寸，与记忆电阻器的特性有关; 它的物理起源是导体中的电荷迁移率。对角矩阵和向量 w = 操作者名{ diag }(vec w)和 vec w 分别是记忆电阻器的内值，值在0到1之间。因此，这个等式需要在内存值上添加额外的约束，以保证可靠性。</nowiki><br />
<br />
==See also==<br />
==相关词条==<br />
{{Columns-list|colwidth=18em|<br />
* [[AI accelerator (computer hardware)|AI accelerator]]<br />
* [[Artificial brain]]<br />
* [[Biomorphic]]<br />
* [[Cognitive computer]]<br />
* [[Computation and Neural Systems]]<br />
* [[Differentiable programming]]<br />
* [[Event camera]]<br />
* [[Neurorobotics]]<br />
* [[Optical flow sensor]]<br />
* [[Physical neural network]]<br />
* [[SpiNNaker]]<br />
* [[SyNAPSE]]<br />
* [[Retinomorphic sensor]]<br />
* [[Vision chip]]<br />
* [[Vision processing unit]]<br />
* [[Zeroth (software)]]<br />
* [[Hardware for artificial intelligence]]<br />
}}<br />
<br />
{{Portal bar|Electronics}}<br />
<br />
==References ==<br />
{{Reflist|40em}}<br />
<br />
==External links==<br />
<!--======================== {{No more links}} ============================<br />
| PLEASE BE CAUTIOUS IN ADDING MORE LINKS TO THIS ARTICLE. Wikipedia |<br />
| is not a collection of links nor should it be used for advertising. |<br />
| |<br />
| Excessive or inappropriate links WILL BE DELETED. |<br />
| See [[Wikipedia:External links]] & [[Wikipedia:Spam]] for details. |<br />
| |<br />
| If there are already plentiful links, please propose additions or |<br />
| replacements on this article's discussion page, or submit your link |<br />
| to the relevant category at the Open Directory Project (dmoz.org) |<br />
| and link back to that category using the {{dmoz}} template. |<br />
======================={{No more links}}=============================--><br />
{{Commons category|Neuromorphic Engineering}}<br />
*[https://web.archive.org/web/20150727034331/http://ine-web.org/workshops/workshops-overview Telluride Neuromorphic Engineering Workshop]<br />
*[https://archive.today/20130115190057/http://capocaccia.ethz.ch/ CapoCaccia Cognitive Neuromorphic Engineering Workshop]<br />
*[https://web.archive.org/web/20190716132350/http://www.ine-web.org/ Institute of Neuromorphic Engineering]<br />
*[http://www.ine-news.org/ INE news site].<br />
*[http://www.frontiersin.org/neuromorphic_engineering Frontiers in Neuromorphic Engineering Journal]<br />
*[http://www.cns.caltech.edu/ Computation and Neural Systems] department at the [[California Institute of Technology]].<br />
*[http://www.humanbrainproject.eu/ Human Brain Project official site]<br />
*[https://www.the-scientist.com/features/building-a-silicon-brain-65738 Building a Silicon Brain:] Computer chips based on biological neurons may help simulate larger and more-complex brain models. May 1, 2019. SANDEEP RAVINDRAN<br />
<br />
<br />
*Telluride Neuromorphic Engineering Workshop<br />
* CapoCaccia Cognitive Neuromorphic Engineering Workshop<br />
* Institute of Neuromorphic Engineering<br />
*INE news site.<br />
* Frontiers in Neuromorphic Engineering Journal<br />
*Computation and Neural Systems department at the California Institute of Technology.<br />
*Human Brain Project official site<br />
*Building a Silicon Brain: Computer chips based on biological neurons may help simulate larger and more-complex brain models. May 1, 2019. SANDEEP RAVINDRAN<br />
<br />
=<nowiki>外部链接</nowiki>= <br />
*碲化物神经形态工程工作室<br />
*CapoCaccia 认知神经形态工程工作室<br />
*神经形态工程研究所<br />
*INE 新闻站点。<br />
*《神经形态工程学前沿》<br />
*加州理工学院计算与神经系统系。<br />
*人脑项目官方网站<br />
*打造硅脑: 基于生物神经元的计算机芯片可能有助于模拟更大、更复杂的大脑模型。2019年5月1日。SANDEEP RAVINDRAN<br />
<br />
{{Differentiable computing}}<br />
{{Authority control}}<br />
<br />
[[Category:Electrical engineering]]<br />
[[Category:Neuroscience]]<br />
[[分类:AI accelerators|*]]<br />
[[Category:Artificial intelligence]]<br />
[[Category:Robotics]]<br />
<br />
Category:Electrical engineering<br />
Category:Neuroscience<br />
*<br />
Category:Artificial intelligence<br />
Category:Robotics<br />
<br />
类别: 电气工程类别: 神经科学 * 类别: 人工智能类别: 机器人<br />
<br />
<noinclude><br />
<br />
<small>This page was moved from [[wikipedia:en:Neuromorphic engineering]]. Its edit history can be viewed at [[神经形态计算/edithistory]]</small></noinclude><br />
<br />
[[Category:待整理页面]]<br />
<br />
<references /></div>Shenky20https://wiki.swarma.org/index.php?title=%E7%A5%9E%E7%BB%8F%E5%BD%A2%E6%80%81%E8%AE%A1%E7%AE%97&diff=30187神经形态计算2022-04-14T08:35:29Z<p>Shenky20：/* 代表性成果 */</p>
<hr />
<div><br />
此词条由神经动力学模型读书会词条梳理志愿者Shenky20翻译审校，未经专家审校，带来阅读不便，请见谅。<br />
<br />
{{Use American English|date = January 2019}}<br />
{{Use mdy dates|date = January 2019}}<br />
<br />
'''Neuromorphic engineering''', also known as '''neuromorphic computing''',<ref name=":3">{{Cite journal | doi = 10.1145/2601069| title = Neuromorphic computing gets ready for the (really) big time| journal = [[Communications of the ACM]]| volume = 57| issue = 6| year = 2014| last1 = Monroe | first1 = D. | pages = 13–15| s2cid = 20051102}}</ref><ref name=":4">{{Cite journal | doi = 10.1088/0957-4484/21/17/175202| pmid = 20368686| title = Nanotube devices based crossbar architecture: Toward neuromorphic computing| journal = Nanotechnology| volume = 21| issue = 17| pages = 175202| year = 2010| last1 = Zhao | first1 = W. S. | last2 = Agnus | first2 = G. | last3 = Derycke | first3 = V. | last4 = Filoramo | first4 = A. | last5 = Bourgoin | first5 = J. -P. | last6 = Gamrat | first6 = C. | bibcode = 2010Nanot..21q5202Z| url = https://zenodo.org/record/3428659}}</ref><ref name="humanbrainproject">{{YouTube|id=6RoiZ90mGfw|title=The Human Brain Project SP 9: Neuromorphic Computing Platform}}</ref> is the use of [[very-large-scale integration]] (VLSI) systems containing electronic [[analog circuit]]s to mimic neuro-biological architectures present in the nervous system. A neuromorphic computer/chip is any device that uses physical artificial neurons (made from silicon) to do computations.<ref name=":5">{{cite journal|last1=Mead|first1=Carver|title=Neuromorphic electronic systems|journal=Proceedings of the IEEE|date=1990|volume=78|issue=10|pages=1629–1636|doi=10.1109/5.58356|url=https://authors.library.caltech.edu/53090/1/00058356.pdf}}</ref><ref name=":2" /> In recent times, the term ''neuromorphic'' has been used to describe analog, digital, mixed-mode analog/digital VLSI, and software systems that implement models of [[neural system]]s (for [[perception]], [[motor control]], or [[multisensory integration]]). The implementation of neuromorphic computing on the hardware level can be realized by oxide-based [[memristor]]s,<ref name="Maan 1–13">{{Cite journal|last1=Maan|first1=A. K.|last2=Jayadevi|first2=D. A.|last3=James|first3=A. P.|date=2016-01-01|title=A Survey of Memristive Threshold Logic Circuits|journal=IEEE Transactions on Neural Networks and Learning Systems|volume=PP|issue=99|pages=1734–1746|doi=10.1109/TNNLS.2016.2547842|pmid=27164608|issn=2162-237X|arxiv=1604.07121|bibcode=2016arXiv160407121M|s2cid=1798273}}</ref> spintronic memories, threshold switches, and [[transistor]]s.<ref name=":6">{{Cite journal|title = Mott Memory and Neuromorphic Devices|journal = Proceedings of the IEEE|date = 2015-08-01|issn = 0018-9219|pages = 1289–1310|volume = 103|issue = 8|doi = 10.1109/JPROC.2015.2431914|first1 = You|last1 = Zhou|first2 = S.|last2 = Ramanathan|s2cid = 11347598|url = https://zenodo.org/record/895565}}</ref><ref name=":2">{{Cite document|title=Neuromorphic Circuits With Neural Modulation Enhancing the Information Content of Neural Signaling {{!}} International Conference on Neuromorphic Systems 2020|language=EN|doi=10.1145/3407197.3407204|s2cid=220794387}}</ref> Training software-based neuromorphic systems of spiking neural networks can be achieved using error backpropagation, e.g., using [[Python (programming language)|Python]] based frameworks such as snnTorch,<ref name=":7">{{cite journal |last1=Eshraghian|first1=Jason K.|last2=Ward|first2=Max|last3=Neftci |first3=Emre|last4=Wang|first4=Xinxin|last5=Lenz|first5=Gregor|last6=Dwivedi|first6=Girish|last7=Bennamoun|first7=Mohammed|last8=Jeong|first8=Doo Seok|last9=Lu|first9=Wei D.|title=Training Spiking Neural Networks Using Lessons from Deep Learning |date=1 October 2021 |arxiv=2109.12894 }}</ref> or using canonical learning rules from the biological learning literature, e.g., using BindsNet.<ref name=":8">{{Cite web | url=https://github.com/Hananel-Hazan/bindsnet | title=Hananel-Hazan/bindsnet: Simulation of spiking neural networks (SNNs) using PyTorch.| date=31 March 2020}}</ref><br />
<br />
<br />
'''<font color="#ff8000">神经形态工程Neuromorphic engineering</font>'''(也称为'''<font color="#ff8000">神经形态计算Neuromorphic computing</font>''')<ref name=":3" /><ref name=":4" /><ref name="humanbrainproject" />是指使用包含电子'''<font color="#ff8000">模拟电路Analog circuit</font>'''的'''<font color="#ff8000">超大规模集成电路Very-large-scale integration</font>'''系统来模拟神经系统中生理结构的研究方法。神经形态计算机或神经形态芯片包括任何使用由硅制成的人造神经元进行计算的设备。<ref name=":5" /><ref name=":2" />近年来，神经形态学（neuromorphic）这个术语被用来描述能够实现'''<font color="#ff8000">神经系统Neural system</font>'''模型功能(如'''<font color="#ff8000">感知Perception</font>'''、'''<font color="#ff8000">运动控制Motor control</font>'''，'''<font color="#ff8000">多感官整合Multisensory integration</font>'''等)的模拟、数字、模拟/数字混合模式超大规模集成电路和软件系统。神经形态计算的硬件实现可以通过基于氧化物的'''<font color="#ff8000">记忆电阻器Memristor</font>'''、<ref name="Maan 1–13" />自旋电子存储器、阈值开关和'''<font color="#ff8000">晶体管Transistor</font>'''来实现。<ref name=":6" /><ref name=":2" />对基于软件的脉冲神经网络系统的训练可以通过误差反向传播机制来实现，例如，使用snnTorch等基于Python的框架，<ref name=":7" />或使用BindsNet等典型的受生物启发的学习模式。<ref name=":8" /><br />
<br />
A key aspect of neuromorphic engineering is understanding how the morphology of individual neurons, circuits, applications, and overall architectures creates desirable computations, affects how information is represented, influences robustness to damage, incorporates learning and development, adapts to local change (plasticity), and facilitates evolutionary change.<br />
<br />
<br />
神经形态工程领域的一个关键问题，就是理解'''<font color="#32CD32">单个神经元形态、神经回路、应用和整体结构</font>'''如何产生理想的计算，如何影响信息的表示和对破坏的鲁棒性，如何整合学习和发展，如何适应局部变化(可塑性) 并促进逐渐发展的变化。<br />
<br />
Neuromorphic engineering is an interdisciplinary subject that takes inspiration from [[biology]], [[physics]], [[mathematics]], [[computer science]], and [[electronic engineering]]<ref name=":2" /> to design artificial neural systems, such as [[Machine vision|vision systems]], head-eye systems, auditory processors, and autonomous robots, whose physical architecture and design principles are based on those of biological nervous systems.<ref name=":9">{{Cite journal | doi = 10.1155/2012/705483| title = Qualitative Functional Decomposition Analysis of Evolved Neuromorphic Flight Controllers| journal = Applied Computational Intelligence and Soft Computing| volume = 2012| pages = 1–21| year = 2012| last1 = Boddhu | first1 = S. K. | last2 = Gallagher | first2 = J. C. | doi-access = free}}</ref> It was developed by [[Carver Mead]]<ref name=":10">{{cite web|title=carver mead website|last1=Mead|first1=Carver|url=http://www.carvermead.caltech.edu/index.html|website=carvermead}}</ref> in the late 1980s.<br />
<br />
<br />
神经形态工程是以'''<font color="#ff8000">生物学Biology</font>'''、'''<font color="#ff8000">物理学Physics</font>'''、'''<font color="#ff8000">数学Mathematics</font>'''、'''<font color="#ff8000">计算机科学Computer science</font>'''和'''<font color="#ff8000">电子工程Electronic engineering<ref name=":2" /></font>'''等学科为基础，设计人工神经系统（如'''<font color="#ff8000">视觉系统Vision systems</font>'''、头眼系统、听觉处理器以及物理结构和设计原则都受启发于生物神经系统的自主机器人）的一门交叉学科。<ref name=":9" />20世纪80年代后期，卡弗·米德极大地推动了神经形态工程领域的发展。<ref name=":10" /><br />
<br />
==Neurological inspiration==<br />
==神经生物学启发==<br />
Neuromorphic engineering is set apart by the inspiration it takes from what we know about the structure and operations of the [[brain]]. Neuromorphic engineering translates what we know about the brain's function into computer systems. Work has mostly focused on replicating the analog nature of [[biological computation]] and the role of [[neuron]]s in [[cognition]].<br />
<br />
<br />
神经形态工程形成的灵感来源于现已知的'''<font color="#ff8000">大脑Brain</font>'''结构及运作机制知识，它将我们对大脑功能的了解用于研究和优化计算机系统。该领域工作主要集中于对'''<font color="#ff8000">生物计算Biological computation</font>'''模拟特性和'''<font color="#ff8000">神经元Neuron</font>'''在'''<font color="#ff8000">认知Cognition</font>'''中所发挥作用的复刻。<br />
<br />
The biological processes of neurons and their [[synapse]]s are dauntingly complex, and thus very difficult to artificially simulate. A key feature of biological brains is that all of the processing in neurons use analog chemical signals. This makes it hard to replicate brains in computers because the current generation of computers is completely digital. However, the characteristics of these parts can be abstracted into mathematical functions that closely capture the essence of the neuron's operations.<br />
<br />
神经元及其'''<font color="#ff8000">突触Synapse</font>'''的生理学过程极其复杂，难以进行人工模拟。神经元在所有处理过程中都使用化学模拟信号，这是大脑的一个关键生理学特征。这个特征大大增加了在计算机中复制大脑的难度，因为目前的计算机是完全数字化的。然而，部分特征可以抽象为数学函数，这些函数能够紧密捕捉神经元运行的本质。<br />
<br />
The goal of neuromorphic computing is not to perfectly mimic the brain and all of its functions, but instead to extract what is known of its structure and operations to be used in a practical computing system. No neuromorphic system will claim nor attempt to reproduce every element of neurons and synapses, but all adhere to the idea that computation is highly [[distributed processing|distributed]] throughout a series of small computing elements analogous to a neuron. While this sentiment is standard, researchers chase this goal with different methods.<ref name=":11">{{Cite journal | doi = 10.1088/1741-2560/13/5/051001| title = Large-scale neuromorphic computing systems| journal = Journal of Neural Engineering| volume = 13| pages = 1–15| year = 2016| last1 = Furber | first1 = Steve| issue = 5| pmid = 27529195| bibcode = 2016JNEng..13e1001F| doi-access = free}}</ref><br />
<br />
神经形态计算的目标不是完美地模拟大脑及其所有功能，而是利用已知的大脑结构和运转机制来研发或优化实际的计算系统。任何神经形态学系统都不会声称或试图复制神经元和突触中的每一个元素，但所有人都一致认可将计算高度'''<font color="#ff8000">分散Distribute</font>'''于一系列类似于神经元的小型计算元素的理念。研究人员用不同的方法来追求这一普遍目标。<ref name=":11" /><br />
<br />
<br />
==Examples==<br />
==代表性成果==<br />
As early as 2006, researchers at [[Georgia Tech]] published a field programmable neural array.<ref name=":12">{{Cite book|title = A field programmable neural array|last1 = Farquhar|first1 = Ethan|date = May 2006|journal = IEEE International Symposium on Circuits and Systems|pages = 4114–4117|last2 = Hasler|first2 = Paul.|doi = 10.1109/ISCAS.2006.1693534|isbn = 978-0-7803-9389-9|s2cid = 206966013}}</ref> This chip was the first in a line of increasingly complex arrays of floating gate transistors that allowed programmability of charge on the gates of [[MOSFET]]s to model the channel-ion characteristics of neurons in the brain and was one of the first cases of a silicon programmable array of neurons.<br />
<br />
早在2006年，佐治亚理工学院的研究人员就研发出了一种现场可编程神经阵列。<ref name=":12" />在此之后，一系列越来越复杂的浮栅晶体管阵列被成功研发出来，这些晶体管阵列可以通过在'''<font color="#ff8000">金属-氧化物半导体效应晶体管MOSFET</font>'''的栅极上编程来模拟大脑中神经元的离子通道特性，这也是以硅为主要材料的可编程神经元阵列的最早成功案例之一。<br />
<br />
In November 2011, a group of [[MIT]] researchers created a computer chip that mimics the analog, ion-based communication in a synapse between two neurons using 400 transistors and standard [[CMOS]] manufacturing techniques.<ref name=":13">{{cite web|title=MIT creates "brain chip"|url=http://www.extremetech.com/extreme/105067-mit-creates-brain-chip|access-date=4 December 2012}}</ref><ref name="Neuromorphic silicon paper">{{cite journal|title=Neuromorphic silicon neurons and large-scale neural networks: challenges and opportunities|doi=10.3389/fnins.2011.00108|pmid=21991244|pmc=3181466|volume=5|pages=108|journal=Frontiers in Neuroscience|year=2011|last1=Poon|first1=Chi-Sang|last2=Zhou|first2=Kuan|doi-access=free}}</ref><br />
<br />
2011年11月，麻省理工学院的一组研究人员研发出一种计算机芯片，该芯片上使用标准的'''<font color="#ff8000">互补金属氧化物半导体CMOS</font>'''制造技术集成了400个晶体管，能够模拟神经元间突触中基于离子的通讯。<ref name=":13" /><ref name="Neuromorphic silicon paper" /><br />
<br />
In June 2012, [[spintronic]] researchers at [[Purdue University]] presented a paper on the design of a neuromorphic chip using [[Spin valve|lateral spin valve]]s and [[memristor]]s. They argue that the architecture works similarly to neurons and can therefore be used to test methods of reproducing the brain's processing. In addition, these chips are significantly more energy-efficient than conventional ones.<ref name="Spin Devices Prop">{{Cite arXiv|title=Proposal For Neuromorphic Hardware Using Spin Devices|eprint=1206.3227|last1=Sharad|first1=Mrigank|last2=Augustine|first2=Charles|last3=Panagopoulos|first3=Georgios|last4=Roy|first4=Kaushik|class=cond-mat.dis-nn|year=2012}}</ref><br />
<br />
2012年6月，普渡大学的'''<font color="#ff8000">自旋电子学Spintronic</font>'''研究人员发表了一篇关于利用'''<font color="#ff8000">侧向自旋阀Lateral spin valves</font>'''和'''<font color="#ff8000">记忆电阻器Memristors</font>'''设计神经形态芯片的论文。他们认为，这种芯片结构的工作原理与神经元相似，因此可以用于大脑运行机制的复刻方法的测试。此外，这些芯片在能耗方面明显优于传统芯片。<ref name="Spin Devices Prop" /><br />
<br />
Research at [[HP Labs]] on Mott memristors has shown that while they can be non-[[Volatile memory|volatile]], the volatile behavior exhibited at temperatures significantly below the [[phase transition]] temperature can be exploited to fabricate a [[neuristor]],<ref name=":0" /> a biologically-inspired device that mimics behavior found in neurons.<ref name=":0">{{Cite journal | doi = 10.1038/nmat3510| pmid = 23241533| title = A scalable neuristor built with Mott memristors| journal = Nature Materials| volume = 12| issue = 2| pages = 114–7| year = 2012| last1 = Pickett | first1 = M. D. | last2 = Medeiros-Ribeiro | first2 = G. | last3 = Williams | first3 = R. S. | bibcode = 2013NatMa..12..114P| s2cid = 16271627| url = https://semanticscholar.org/paper/b6ba6f496ace2b947f111059663e76bb60e9efeb}}</ref> In September 2013, they presented models and simulations that show how the spiking behavior of these neuristors can be used to form the components required for a [[Turing machine]].<ref name=":14">{{cite journal|doi=10.1088/0957-4484/24/38/384002|title=Phase transitions enable computational universality in neuristor-based cellular automata|author1=Matthew D Pickett|author2=R Stanley Williams|name-list-style=amp|date=September 2013|publisher=IOP Publishing Ltd|journal=Nanotechnology|volume=24|issue=38|pmid=23999059|bibcode=2013Nanot..24L4002P|at=384002}}</ref><br />
<br />
'''<font color="#ff8000">惠普实验室HP labs</font>'''在莫特记忆电阻器上的研究表明，尽管它们可以是非'''<font color="#ff8000">易失性Volatile</font>'''的，但是在'''<font color="#ff8000">相变Phase transition</font>'''温度以下时表现出的易失性行为可以被用来制造'''<font color="#ff8000">类神经元电阻器Neuristor</font>'''(一种生物学启发的模仿神经元行为的硬件)<ref name=":0" />。2013年9月，他们通过模型和仿真展示了这些类神经元电阻器的脉冲行为如何产生'''<font color="#ff8000">图灵机Turing machine</font>'''的所需元素。<ref name=":14" /><br />
<br />
[[Neurogrid]], built by ''Brains in Silicon'' at [[Stanford University]],<ref name=":15">{{cite journal|last1=Boahen|first1=Kwabena|title=Neurogrid: A Mixed-Analog-Digital Multichip System for Large-Scale Neural Simulations|journal=Proceedings of the IEEE|date=24 April 2014|volume=102|issue=5|pages=699–716|doi=10.1109/JPROC.2014.2313565|s2cid=17176371}}</ref> is an example of hardware designed using neuromorphic engineering principles. The circuit board is composed of 16 custom-designed chips, referred to as NeuroCores. Each NeuroCore's analog circuitry is designed to emulate neural elements for 65536 neurons, maximizing energy efficiency. The emulated neurons are connected using digital circuitry designed to maximize spiking throughput.<ref name=":16">{{cite journal|doi=10.1038/503022a|pmid = 24201264|title = Neuroelectronics: Smart connections|journal = Nature|volume = 503|issue = 7474|pages = 22–4|year = 2013|last1 = Waldrop|first1 = M. Mitchell|bibcode = 2013Natur.503...22W|doi-access = free}}</ref><ref name=":17">{{cite journal|doi=10.1109/JPROC.2014.2313565|title = Neurogrid: A Mixed-Analog-Digital Multichip System for Large-Scale Neural Simulations|journal = Proceedings of the IEEE|volume = 102|issue = 5|pages = 699–716|year = 2014|last1 = Benjamin|first1 = Ben Varkey|last2 = Peiran Gao|last3 = McQuinn|first3 = Emmett|last4 = Choudhary|first4 = Swadesh|last5 = Chandrasekaran|first5 = Anand R.|last6 = Bussat|first6 = Jean-Marie|last7 = Alvarez-Icaza|first7 = Rodrigo|last8 = Arthur|first8 = John V.|last9 = Merolla|first9 = Paul A.|last10 = Boahen|first10 = Kwabena|s2cid = 17176371}}</ref><br />
<br />
'''<font color="#ff8000">神经栅格Neurogrid<ref name=":15" /></font>'''是由斯坦福大学Brains in Silicon公司研发的、使用神经形态工程原理设计的硬件。该电路板由16个定制设计的芯片组成(NeuroCores)。在设计中，每个NeuroCore芯片的模拟电路对65536个神经元的神经元素进行模拟，以最大限度地提高能量效率。模拟出的神经元通过设计的数字电路连接，以最大化脉冲吞吐量。<ref name=":16" /><ref name=":17" /><br />
<br />
A research project with implications for neuromorphic engineering is the [[Human Brain Project]] that is attempting to simulate a complete human brain in a supercomputer using biological data. It is made up of a group of researchers in neuroscience, medicine, and computing.<ref name=":18">{{cite web|title=Involved Organizations|url=http://www.humanbrainproject.eu/partners.html|access-date=22 February 2013|url-status=dead|archive-url=https://web.archive.org/web/20130302142627/http://www.humanbrainproject.eu/partners.html|archive-date=2 March 2013}}</ref> [[Henry Markram]], the project's co-director, has stated that the project proposes to establish a foundation to explore and understand the brain and its diseases, and to use that knowledge to build new computing technologies. The three primary goals of the project are to better understand how the pieces of the brain fit and work together, to understand how to objectively diagnose and treat brain diseases, and to use the understanding of the human brain to develop neuromorphic computers. That the simulation of a complete human brain will require a supercomputer a thousand times more powerful than today's encourages the current focus on neuromorphic computers.<ref name=":19">{{cite web|title=Human Brain Project|url=http://www.humanbrainproject.eu|access-date=22 February 2013}}</ref> $1.3 billion has been allocated to the project by The [[European Commission]].<ref name=":20">{{cite web|title=The Human Brain Project and Recruiting More Cyberwarriors|url=http://www.marketplace.org/topics/tech/human-brain-project-and-recruiting-more-cyberwarriors|access-date=22 February 2013|date=January 29, 2013}}</ref><br />
<br />
'''<font color="#ff8000">人类大脑计划Human Brain Project</font>'''对神经形态工程具有较大影响，其主要任务是尝试用生物数据在超级计算机中模拟完整的人脑。人类大脑计划由神经科学、医学和计算机科学背景的研究人员组成。<ref name=":18" />该项目的联合主管亨利•马克拉姆(Henry Markram)表示，人类大脑计划的目的是建立一个探索和了解脑科学和脑疾病知识的基础，并利用这些知识来构建更先进的计算机技术。这个项目的三个主要目标分别是: 更好地理解大脑的各个部分是如何相互配合协同工作的; 理解如何客观地诊断和治疗脑部疾病; 以及利用对人类大脑的理解来开发神经形态计算机。模拟一个完整的人类大脑需要一台比现在强大一千倍的超级计算机，这不断激发着对神经形态计算机领域的研究兴趣。<ref name=":19" />欧盟委员会已经向人类大脑计划拨款13亿美元。<ref name=":20" /><br />
<br />
Other research with implications for neuromorphic engineering involves the [[BRAIN Initiative]]<ref name="economist">[https://www.economist.com/news/science-and-technology/21582495-computers-will-help-people-understand-brains-better-and-understanding-brains Neuromorphic computing: The machine of a new soul], The Economist, 2013-08-03</ref> and the [[TrueNorth]] chip from [[IBM]].<ref name=":21">{{cite journal|last1=Modha|first1=Dharmendra|title=A million spiking-neuron integrated circuit with a scalable communication network and interface|journal=Science|date=Aug 2014|volume=345|issue=6197|pages=668–673|doi=10.1126/science.1254642|pmid=25104385|bibcode=2014Sci...345..668M|s2cid=12706847}}</ref> Neuromorphic devices have also been demonstrated using nanocrystals, nanowires, and conducting polymers.<ref name=":22">{{Cite web|url=http://jessamynfairfield.com/wp-content/uploads/2017/03/PWMar17Fairfield.pdf|title=Smarter Machines|last=Fairfield|first=Jessamyn|date=March 1, 2017}}</ref><br />
<br />
其他与神经形态工程有关的研究还包括'''<font color="#ff8000">脑计划BRAIN initiative</font>'''，<ref name="economist" />和IBM研发的'''<font color="#ff8000">TrueNorth</font>'''芯片。<ref name=":21" />使用纳米晶体、纳米线和导电聚合物也能够用于制造神经形态学硬件。<ref name=":22" /><br />
<br />
[[Intel]] unveiled its neuromorphic research chip, called “[[Intel Loihi|Loihi]]”, in October 2017. The chip uses an asynchronous [[spiking neural network]] (SNN) to implement adaptive self-modifying event-driven fine-grained parallel computations used to implement learning and inference with high efficiency.<ref name=":23">{{cite journal |last1=Davies |first1=Mike |title=Loihi: A Neuromorphic Manycore Processor with On-Chip Learning |journal=IEEE Micro |date=January 16, 2018 |volume=38 |issue=1 |pages=82–99 |display-authors=etal|doi=10.1109/MM.2018.112130359 |s2cid=3608458 }}</ref><ref name="Morris2017">{{cite web |last1=Morris |first1=John |title=Why Intel built a neuromorphic chip |url=https://www.zdnet.com/article/why-intel-built-a-neuromorphic-chip/ |website=ZDNet |access-date=17 August 2018 |language=en}}</ref><br />
<br />
2017年10月，英特尔发布了神经形态芯片'''<font color="#ff8000">Loihi</font>'''。该芯片采用异步'''<font color="#ff8000">脉冲神经网络Spiking neural network</font>'''实现了自适应、自修改、事件驱动的细粒度并行计算，实现了高效的学习和推理。<ref name=":23" /><ref name="Morris2017" /><br />
<br />
[[IMEC]], a Belgium-based nanoelectronics research center, demonstrated the world's first self-learning neuromorphic chip. The brain-inspired chip, based on OxRAM technology, has the capability of self-learning and has been demonstrated to have the ability to compose music.<ref name=":24">{{cite web |title=Imec demonstrates self-learning neuromorphic chip that composes music |url=https://www.imec-int.com/en/articles/imec-demonstrates-self-learning-neuromorphic-chip-that-composes-music |website=IMEC International |access-date=1 October 2019}}</ref> IMEC released the 30-second tune composed by the prototype. The chip was sequentially loaded with songs in the same time signature and style. The songs were old Belgian and French flute minuets, from which the chip learned the rules at play and then applied them.<ref name=":25">{{cite web|last1=Bourzac|first1=Katherine|title=A Neuromorphic Chip That Makes Music|url=https://spectrum.ieee.org/a-neuromorphic-chip-that-makes-music|url-status=live|access-date=1 October 2019|website=IEEE Spectrum|date=May 23, 2017}}</ref><br />
<br />
比利时的'''<font color="#ff8000">微电子研究中心IMEC</font>'''研发了世界上首个自学习神经形态芯片。这种基于'''<font color="#32CD32"> OxRAM(filamentary-oxide-based resistive memory technology)</font>''' 技术的大脑启发芯片具有自学习能力，并且已被证明具有创作音乐的能力。<ref name=":24" />IMEC发布了由芯片原型机谱写的30秒旋律。向芯片加载一系列特征、风格统一的歌曲（古代比利时和法国长笛小步舞曲），芯片从中学习相关规则并将其应用于创作。<ref name=":25" /><br />
<br />
[[Blue Brain Project|The Blue Brain Project]], led by Henry Markram, aims to build biologically detailed digital reconstructions and simulations of the mouse brain. The Blue Brain Project has created in silico models of rodent brains, while attempting to replicate as many details about its biology as possible. The supercomputer-based simulations offer new perspectives on understanding the structure and functions of the brain.<br />
<br />
由亨利·马克拉姆领导的'''<font color="#ff8000">蓝脑计划The Blue Brain Project</font>'''旨在建立小鼠大脑生理学细节的数字重建和模拟。蓝脑计划已经建立了啮齿动物大脑的电子模型，同时进行着尽可能多地复制其生理学细节的尝试。基于超级计算机的模拟为理解大脑的结构和功能提供了新的视角。<br />
<br />
The European Union funded a series of projects at the University of Heidelberg, which led to the development of <br />
[[BrainScaleS]] (brain-inspired multiscale computation in neuromorphic hybrid systems), a hybrid analog [[neuromorphic]] supercomputer located at Heidelberg University, Germany. It was developed as part of the [[Human Brain Project]] neuromorphic computing platform and is the complement to the [[SpiNNaker]] supercomputer (which is based on digital technology). The architecture used in BrainScaleS mimics biological neurons and their connections on a physical level; additionally, since the components are made of silicon, these model neurons operate on average 864 times (24 hours of real time is 100 seconds in the machine simulation) that of their biological counterparts.<ref name=":26">{{Cite web|date=2016-03-21|title=Beyond von Neumann, Neuromorphic Computing Steadily Advances|url=https://www.hpcwire.com/2016/03/21/lacking-breakthrough-neuromorphic-computing-steadily-advance/|access-date=2021-10-08|website=HPCwire|language=en-US}}</ref><br />
<br />
欧盟资助了海德堡大学一系列促进BrainScaleS(神经形态混合系统中受大脑启发的多尺度计算)发展的项目，这是一台位于德国海德堡大学的混合模拟'''<font color="#ff8000">神经形态Neuromorphic</font>'''超级计算机。它是作为人类大脑计划中神经形态计算平台的一部分而开发的，是'''<font color="#ff8000">SpiNNaker</font>'''超级计算机(基于数字技术)的补充。BrainScaleS中使用的体系架构模拟了生物神经元及其在物理层面上的连接；此外，由于这些组件是由硅制成的，这些模型神经元平均运行速度是生物神经元的864倍，这意味着在机器模拟中，24小时的实时时间仅为100秒。<ref name=":26" /><br />
<br />
===Neuromorphic sensors===<br />
===神经形态传感器===<br />
The concept of neuromorphic systems can be extended to sensors (not just to computation). An example of this applied to detecting [[light]] is the [[retinomorphic sensor]] or, when employed in an array, the [[event camera]].<br />
<br />
神经形态系统的概念可以扩展到传感器(而不仅仅是计算单元)。一个用于检测光线的例子是'''<font color="#ff8000">类视网膜传感器Retinomorphic sensor</font>'''，或者'''<font color="#ff8000">事件摄像机Event camera</font>阵列。'''<br />
<br />
<br />
==Ethical considerations ==<br />
==伦理问题==<br />
While the interdisciplinary concept of neuromorphic engineering is relatively new, many of the same ethical considerations apply to neuromorphic systems as apply to [[human-like machines]] and [[artificial intelligence]] in general. However, the fact that neuromorphic systems are designed to mimic a [[human brain]] gives rise to unique ethical questions surrounding their usage.<br />
<br />
While the interdisciplinary concept of neuromorphic engineering is relatively new, many of the same ethical considerations apply to neuromorphic systems as apply to human-like machines and artificial intelligence in general. However, the fact that neuromorphic systems are designed to mimic a human brain gives rise to unique ethical questions surrounding their usage.<br />
<br />
虽然神经形态工程的跨学科概念相对较新，但许多同样的伦理考虑适用于神经形态系统，就像适用于类人机器和一般人工智能一样。然而，神经形态系统是为了模仿人类大脑而设计的这一事实引起了围绕其使用的独特的伦理问题。<br />
<br />
However, the practical debate is that neuromorphic hardware as well as artificial "neural networks" are immensely simplified models of how the brain operates or processes information at a much lower [[complex system|complexity in terms of size and functional technology]] and a much more regular structure in terms of [[brain connectivity|connectivity]]. Comparing [[neuromorphic chip]]s to the brain is a very crude comparison similar to comparing a plane to a bird just because they both have wings and a tail. The fact is that neural cognitive systems are many orders of magnitude more [[Energy efficiency (physics)|energy-]] and compute-efficient than current state-of-the-art AI and neuromorphic engineering is an attempt to narrow this gap by inspiring from the brain's mechanism just like many engineering designs have [[bioengineering|bio-inspired features]].<br />
<br />
However, the practical debate is that neuromorphic hardware as well as artificial "neural networks" are immensely simplified models of how the brain operates or processes information at a much lower complexity in terms of size and functional technology and a much more regular structure in terms of connectivity. Comparing neuromorphic chips to the brain is a very crude comparison similar to comparing a plane to a bird just because they both have wings and a tail. The fact is that neural cognitive systems are many orders of magnitude more energy- and compute-efficient than current state-of-the-art AI and neuromorphic engineering is an attempt to narrow this gap by inspiring from the brain's mechanism just like many engineering designs have bio-inspired features.<br />
<br />
然而，实际上的争论是，神经形态硬件和人工“神经网络”是大脑如何运作或处理信息的极其简化的模型，在大小和功能技术方面的复杂性要低得多，在连接方面的结构则更加规则。将神经形态芯片与大脑进行比较是一种非常粗糙的比较，类似于仅仅因为一架飞机有翅膀和一条尾巴就将它与一只鸟进行比较。事实上，神经认知系统比当前最先进的人工智能具有更多的能量和计算效率，而神经形态工程是一种通过从大脑机制中激发灵感来缩小这种差距的尝试，就像许多工程设计具有生物启发的特征一样。数量级。<br />
<br />
===Democratic concerns===<br />
===公众担忧 ===<br />
Significant ethical limitations may be placed on neuromorphic engineering due to public perception.<ref>{{Cite report|url=https://ai100.stanford.edu/sites/g/files/sbiybj9861/f/ai_100_report_0831fnl.pdf|title=Artificial Intelligence and Life in 2030|author=2015 Study Panel|date=September 2016|work=One Hundred Year Study on Artificial Intelligence (AI100)|publisher=Stanford University}}</ref> Special [[Eurobarometer]] 382: Public Attitudes Towards Robots, a survey conducted by the European Commission, found that 60% of [[European Union]] citizens wanted a ban of robots in the care of children, the elderly, or the disabled. Furthermore, 34% were in favor of a ban on robots in education, 27% in healthcare, and 20% in leisure. The European Commission classifies these areas as notably “human.” The report cites increased public concern with robots that are able to mimic or replicate human functions. Neuromorphic engineering, by definition, is designed to replicate the function of the human brain.<ref name=":1">{{Cite web|url=http://ec.europa.eu/commfrontoffice/publicopinion/archives/ebs/ebs_382_en.pdf|title=Special Eurobarometer 382: Public Attitudes Towards Robots|last=European Commission|date=September 2012|website=European Commission}}</ref><br />
<br />
Significant ethical limitations may be placed on neuromorphic engineering due to public perception. Special Eurobarometer 382: Public Attitudes Towards Robots, a survey conducted by the European Commission, found that 60% of European Union citizens wanted a ban of robots in the care of children, the elderly, or the disabled. Furthermore, 34% were in favor of a ban on robots in education, 27% in healthcare, and 20% in leisure. The European Commission classifies these areas as notably “human.” The report cites increased public concern with robots that are able to mimic or replicate human functions. Neuromorphic engineering, by definition, is designed to replicate the function of the human brain.<br />
<br />
由于公众的认知，神经形态工程学可能会受到严重的伦理限制。欧盟委员会进行的一项调查发现，60% 的欧盟公民希望禁止照顾儿童、老人或残疾人的机器人。此外，34% 的人支持禁止机器人用于教育，27% 的人支持医疗保健，20% 的人支持休闲。欧盟委员会将这些地区明显归类为“人类”报告指出，公众越来越关注能够模仿或复制人类功能的机器人。神经形态工程，顾名思义，是为了复制人脑的功能而设计的。<br />
<br />
The democratic concerns surrounding neuromorphic engineering are likely to become even more profound in the future. The European Commission found that EU citizens between the ages of 15 and 24 are more likely to think of robots as human-like (as opposed to instrument-like) than EU citizens over the age of 55. When presented an image of a robot that had been defined as human-like, 75% of EU citizens aged 15–24 said it corresponded with the idea they had of robots while only 57% of EU citizens over the age of 55 responded the same way. The human-like nature of neuromorphic systems, therefore, could place them in the categories of robots many EU citizens would like to see banned in the future.<ref name=":1" /><br />
<br />
The democratic concerns surrounding neuromorphic engineering are likely to become even more profound in the future. The European Commission found that EU citizens between the ages of 15 and 24 are more likely to think of robots as human-like (as opposed to instrument-like) than EU citizens over the age of 55. When presented an image of a robot that had been defined as human-like, 75% of EU citizens aged 15–24 said it corresponded with the idea they had of robots while only 57% of EU citizens over the age of 55 responded the same way. The human-like nature of neuromorphic systems, therefore, could place them in the categories of robots many EU citizens would like to see banned in the future.<br />
<br />
围绕神经形态工程的民主关注可能在未来变得更加深刻。欧盟委员会(European Commission)发现，15至24岁的欧盟公民比55岁以上的欧盟公民更有可能认为机器人像人(而不是像仪器)。当看到一张被定义为“类人”的机器人图片时，年龄在15岁至24岁之间的欧盟公民中有75% 的人表示，这与他们对机器人的想法相符，而55岁以上的欧盟公民中只有57% 的人有同样的反应。因此，类似人类的神经形态系统，可以把它们归入许多欧盟公民希望在未来禁止使用的机器人类别。<br />
<br />
=== Personhood===<br />
===人格问题===<br />
As neuromorphic systems have become increasingly advanced, some scholars{{who|date=August 2021}} have advocated for granting [[personhood]] rights to these systems. If the brain is what grants humans their personhood, to what extent does a neuromorphic system have to mimic the human brain to be granted personhood rights? Critics of technology development in the [[Human Brain Project]], which aims to advance brain-inspired computing, have argued that advancement in neuromorphic computing could lead to machine consciousness or personhood.<ref>{{Cite journal|last=Aicardi|first=Christine|date=September 2018|title=Accompanying technology development in the Human Brain Project: From foresight to ethics management|journal=Futures|volume=102|pages=114–124|doi=10.1016/j.futures.2018.01.005|doi-access=free}}</ref> If these systems are to be treated as people, critics argue, then many tasks humans perform using neuromorphic systems, including the act of termination of neuromorphic systems, may be morally impermissible as these acts would violate the autonomy of the neuromorphic systems.<ref>{{Cite journal|last=Lim|first=Daniel|date=2014-06-01|title=Brain simulation and personhood: a concern with the Human Brain Project|journal=Ethics and Information Technology|language=en|volume=16|issue=2|pages=77–89|doi=10.1007/s10676-013-9330-5|s2cid=17415814|issn=1572-8439}}</ref><br />
<br />
As neuromorphic systems have become increasingly advanced, some scholars have advocated for granting personhood rights to these systems. If the brain is what grants humans their personhood, to what extent does a neuromorphic system have to mimic the human brain to be granted personhood rights? Critics of technology development in the Human Brain Project, which aims to advance brain-inspired computing, have argued that advancement in neuromorphic computing could lead to machine consciousness or personhood. If these systems are to be treated as people, critics argue, then many tasks humans perform using neuromorphic systems, including the act of termination of neuromorphic systems, may be morally impermissible as these acts would violate the autonomy of the neuromorphic systems.<br />
<br />
随着神经形态系统的日益发展，一些学者主张赋予这些系统人格权。如果是大脑赋予了人类人格，那么神经形态系统在多大程度上必须模仿人类大脑才能被赋予人格权利？“人脑计划”旨在推进以大脑为灵感的计算机技术发展，该计划的批评者认为，神经形态计算机技术的进步可能导致机器意识或人格的形成。批评家认为，如果这些系统被当作人来对待，那么人类使用神经形态系统执行的许多任务，包括终止神经形态系统的行为，可能在道德上是不允许的，因为这些行为将违反神经形态系统的自主性。<br />
<br />
==Dual use (military applications)==<br />
==军民两用技术==<br />
The [[Joint Artificial Intelligence Center]], a branch of the U.S. military, is a center dedicated to the procurement and implementation of AI software and neuromorphic hardware for combat use. Specific applications include smart headsets/goggles and robots. JAIC intends to rely heavily on neuromorphic technology to connect "every fighter every shooter" within a network of neuromorphic-enabled units.<br />
<br />
The Joint Artificial Intelligence Center, a branch of the U.S. military, is a center dedicated to the procurement and implementation of AI software and neuromorphic hardware for combat use. Specific applications include smart headsets/goggles and robots. JAIC intends to rely heavily on neuromorphic technology to connect "every fighter every shooter" within a network of neuromorphic-enabled units.<br />
<br />
军民两用联合人工智能中心是美国军队的一个分支，专门从事采购和实施用于战斗的人工智能软件和神经形态硬件。具体应用包括智能耳机/护目镜和机器人。JAIC 打算严重依赖神经形态技术来连接神经形态单位网络中的“每个战士每个射手”。<br />
<br />
==Legal considerations==<br />
==法律问题==<br />
Skeptics have argued that there is no way to apply the electronic personhood, the concept of personhood that would apply to neuromorphic technology, legally. In a letter signed by 285 experts in law, robotics, medicine, and ethics opposing a European Commission proposal to recognize “smart robots” as legal persons, the authors write, “A legal status for a robot can’t derive from the [[Natural person|Natural Person]] model, since the robot would then hold [[human rights]], such as the right to dignity, the right to its integrity, the right to remuneration or the right to citizenship, thus directly confronting the Human rights. This would be in contradiction with the [[Charter of Fundamental Rights of the European Union]] and the [[Convention for the Protection of Human Rights and Fundamental Freedoms]].”<ref>{{Cite web|url=http://www.robotics-openletter.eu/|title=Robotics Openletter {{!}} Open letter to the European Commission|language=fr-FR|access-date=2019-05-10}}</ref><br />
<br />
Skeptics have argued that there is no way to apply the electronic personhood, the concept of personhood that would apply to neuromorphic technology, legally. In a letter signed by 285 experts in law, robotics, medicine, and ethics opposing a European Commission proposal to recognize “smart robots” as legal persons, the authors write, “A legal status for a robot can’t derive from the Natural Person model, since the robot would then hold human rights, such as the right to dignity, the right to its integrity, the right to remuneration or the right to citizenship, thus directly confronting the Human rights. This would be in contradiction with the Charter of Fundamental Rights of the European Union and the Convention for the Protection of Human Rights and Fundamental Freedoms.”<br />
<br />
怀疑论者认为，在法律上没有办法应用电子人格，这个人格概念将适用于神经形态技术。在一封由285名法律、机器人、医学和伦理学专家签名的信中，作者们反对欧盟委员会承认“智能机器人”为法人的提议，他们写道，“机器人的法律地位不能从自然人模型中推导出来，因为机器人将拥有人权，如尊严权、完整权、报酬权或公民权，从而直接面对人权。这将有悖于《欧洲联盟基本权利宪章和《保护人权和基本自由公约》。”<br />
<br />
===Ownership and property rights===<br />
===所有权及财产权问题===<br />
There is significant legal debate around property rights and artificial intelligence. In ''Acohs Pty Ltd v. Ucorp Pty Ltd'', Justice Christopher Jessup of the [[Federal Court of Australia]] found that the [[source code]] for [[Material safety data sheets|Material Safety Data Sheets]] could not be [[Copyright law of Australia|copyrighted]] as it was generated by a [[software interface]] rather than a human author.<ref>{{Cite web|url=http://www.lavan.com.au/advice/intellectual_property_technology/copyright_in_source_code_and_digital_products|title=Copyright in source code and digital products|last=Lavan|website=Lavan|language=en|access-date=2019-05-10}}</ref> The same question may apply to neuromorphic systems: if a neuromorphic system successfully mimics a human brain and produces a piece of original work, who, if anyone, should be able to claim ownership of the work?<ref>{{cite journal |last1=Eshraghian|first1=Jason K. |title=Human Ownership of Artificial Creativity |journal=Nature Machine Intelligence |date=9 March 2020 |volume=2 |pages=157–160 |doi=10.1038/s42256-020-0161-x}}</ref><br />
<br />
There is significant legal debate around property rights and artificial intelligence. In Acohs Pty Ltd v. Ucorp Pty Ltd, Justice Christopher Jessup of the Federal Court of Australia found that the source code for Material Safety Data Sheets could not be copyrighted as it was generated by a software interface rather than a human author. The same question may apply to neuromorphic systems: if a neuromorphic system successfully mimics a human brain and produces a piece of original work, who, if anyone, should be able to claim ownership of the work?<br />
<br />
围绕财产权和人工智能有着重大的法律争论。在 Acohs Pty Ltd 诉 Ucorp Pty Ltd 一案中，澳大利亚联邦法院的克里斯托弗 · 杰瑟普法官发现，材料安全数据表的源代码不能受版权保护，因为它是由软件界面而不是人工作者生成的。同样的问题可能也适用于神经形态系统: 如果一个神经形态系统成功地模仿了人类的大脑并产生了一部原创作品，那么谁，如果有人，应该声称拥有这部作品的所有权？<br />
<br />
==Neuromemristive systems==<br />
==神经记忆电阻系统==<br />
Neuromemristive systems are a subclass of neuromorphic computing systems that focus on the use of [[memristors]] to implement [[neuroplasticity]]. While neuromorphic engineering focuses on mimicking biological behavior, neuromemristive systems focus on abstraction.<ref>{{Cite web|url=https://digitalops.sandia.gov/Mediasite/Play/a10cf6ceb55d47608bb8326dd00e46611d|title=002.08 N.I.C.E. Workshop 2014: Towards Intelligent Computing with Neuromemristive Circuits and Systems - Feb. 2014|website=digitalops.sandia.gov|access-date=2019-08-26}}</ref> For example, a neuromemristive system may replace the details of a [[Cerebral cortex|cortical]] microcircuit's behavior with an abstract neural network model.<ref>C. Merkel and D. Kudithipudi, "Neuromemristive extreme learning machines for pattern classification," ISVLSI, 2014.</ref><br />
<br />
Neuromemristive systems are a subclass of neuromorphic computing systems that focus on the use of memristors to implement neuroplasticity. While neuromorphic engineering focuses on mimicking biological behavior, neuromemristive systems focus on abstraction. For example, a neuromemristive system may replace the details of a cortical microcircuit's behavior with an abstract neural network model.C. Merkel and D. Kudithipudi, "Neuromemristive extreme learning machines for pattern classification," ISVLSI, 2014.<br />
<br />
神经记忆电阻系统是神经形态计算系统的一个亚类，主要研究利用记忆电阻器实现神经可塑性。神经形态工程的重点是模拟生物行为，而神经记忆电阻系统的重点是提取。例如，一个神经记忆系统可以用一个抽象的神经网络模型取代皮层微电路的行为细节。默克尔和 d. Kudithipudi，“用于模式分类的神经记忆极端学习机器”，ISVLSI，2014。<br />
<br />
There exist several neuron inspired threshold logic functions<ref name="Maan 1–13" /> implemented with memristors that have applications in high level [[pattern recognition]] applications. Some of the applications reported recently include [[speech recognition]],<ref>{{Cite journal|title = Memristor pattern recogniser: isolated speech word recognition|journal = Electronics Letters|pages = 1370–1372|volume = 51|issue = 17|doi = 10.1049/el.2015.1428|first1 = A.K.|last1 = Maan|first2 = A.P.|last2 = James|first3 = S.|last3 = Dimitrijev|year = 2015|bibcode = 2015ElL....51.1370M|hdl = 10072/140989|s2cid = 61454815|url = https://semanticscholar.org/paper/48d3ab11ec6e213b62f11eedcfb7b7febb058674|hdl-access = free}}</ref> [[face recognition]]<ref>{{Cite journal|title = Memristive Threshold Logic Face Recognition|journal = Procedia Computer Science|date = 2014-01-01|pages = 98–103|volume = 41|series = 5th Annual International Conference on Biologically Inspired Cognitive Architectures, 2014 BICA|doi = 10.1016/j.procs.2014.11.090|first1 = Akshay Kumar|last1 = Maan|first2 = Dinesh S.|last2 = Kumar|first3 = Alex Pappachen|last3 = James|doi-access = free}}</ref> and [[object recognition]].<ref>{{Cite journal|title = Memristive Threshold Logic Circuit Design of Fast Moving Object Detection|journal = IEEE Transactions on Very Large Scale Integration (VLSI) Systems|date = 2015-10-01|issn = 1063-8210|pages = 2337–2341|volume = 23|issue = 10|doi = 10.1109/TVLSI.2014.2359801|first1 = A.K.|last1 = Maan|first2 = D.S.|last2 = Kumar|first3 = S.|last3 = Sugathan|first4 = A.P.|last4 = James|arxiv = 1410.1267|s2cid = 9647290}}</ref> They also find applications in replacing conventional digital logic gates.<ref>{{Cite journal|title = Resistive Threshold Logic|journal = IEEE Transactions on Very Large Scale Integration (VLSI) Systems|date = 2014-01-01|issn = 1063-8210|pages = 190–195|volume = 22|issue = 1|doi = 10.1109/TVLSI.2012.2232946|first1 = A.P.|last1 = James|first2 = L.R.V.J.|last2 = Francis|first3 = D.S.|last3 = Kumar|arxiv = 1308.0090|s2cid = 7357110}}</ref><ref>{{Cite journal|title = Threshold Logic Computing: Memristive-CMOS Circuits for Fast Fourier Transform and Vedic Multiplication|journal = IEEE Transactions on Very Large Scale Integration (VLSI) Systems|date = 2015-11-01|issn = 1063-8210|pages = 2690–2694|volume = 23|issue = 11|doi = 10.1109/TVLSI.2014.2371857|first1 = A.P.|last1 = James|first2 = D.S.|last2 = Kumar|first3 = A.|last3 = Ajayan|arxiv = 1411.5255|s2cid = 6076956}}</ref><br />
<br />
There exist several neuron inspired threshold logic functions implemented with memristors that have applications in high level pattern recognition applications. Some of the applications reported recently include speech recognition, face recognition and object recognition. They also find applications in replacing conventional digital logic gates.<br />
<br />
利用记忆电阻器实现的神经元启发阈值逻辑函数在高级模式识别中有着广泛的应用。最近报道的一些应用包括语音识别、人脸识别和物体识别。它们还可以用来取代传统的数字逻辑门。<br />
<br />
For ideal passive memristive circuits there is an exact equation (Caravelli-Traversa-[[Di Ventra]] equation) for the internal memory of the circuit:<ref>{{cite journal |last=Caravelli |display-authors=etal|arxiv=1608.08651 |title=The complex dynamics of memristive circuits: analytical results and universal slow relaxation |year=2017 |doi=10.1103/PhysRevE.95.022140 |pmid= 28297937 |volume=95 |issue= 2 |pages= 022140 |journal=Physical Review E|bibcode=2017PhRvE..95b2140C |s2cid=6758362}}</ref><br />
<br />
For ideal passive memristive circuits there is an exact equation (Caravelli-Traversa-Di Ventra equation) for the internal memory of the circuit:<br />
<br />
对于理想的无源记忆电路，电路的内部记忆有一个精确的方程(Caravelli-Traversa-Di Ventra 方程) :<br />
<br />
:<math> \frac{d}{dt} \vec{W} = \alpha \vec{W}-\frac{1}{\beta} (I+\xi \Omega W)^{-1} \Omega \vec S </math><br />
<br />
:\frac{d}{dt} \vec{W} = \alpha \vec{W}-\frac{1}{\beta} (I+\xi \Omega W)^{-1} \Omega \vec S<br />
<br />
:\frac{d}{dt} \vec{W} = \alpha \vec{W}-\frac{1}{\beta} (I+\xi \Omega W)^{-1} \Omega \vec S<br />
<br />
as a function of the properties of the physical memristive network and the external sources. In the equation above, <math>\alpha</math> is the "forgetting" time scale constant, <math> \xi=r-1</math> and <math>r=\frac{R_\text{off}}{R_\text{on}}</math> is the ratio of ''off'' and ''on'' values of the limit resistances of the memristors, <math> \vec S </math> is the vector of the sources of the circuit and <math>\Omega</math> is a projector on the fundamental loops of the circuit. The constant <math>\beta</math> has the dimension of a voltage and is associated to the properties of the [[memristor]]; its physical origin is the charge mobility in the conductor. The diagonal matrix and vector <math>W=\operatorname{diag}(\vec W)</math> and <math>\vec W</math> respectively, are instead the internal value of the memristors, with values between 0 and 1. This equation thus requires adding extra constraints on the memory values in order to be reliable.<br />
<br />
<nowiki>as a function of the properties of the physical memristive network and the external sources. In the equation above, \alpha is the "forgetting" time scale constant, \xi=r-1 and r=\frac{R_\text{off}}{R_\text{on}} is the ratio of off and on values of the limit resistances of the memristors, \vec S is the vector of the sources of the circuit and \Omega is a projector on the fundamental loops of the circuit. The constant \beta has the dimension of a voltage and is associated to the properties of the memristor; its physical origin is the charge mobility in the conductor. The diagonal matrix and vector W=\operatorname{diag}(\vec W) and \vec W respectively, are instead the internal value of the memristors, with values between 0 and 1. This equation thus requires adding extra constraints on the memory values in order to be reliable.</nowiki><br />
<br />
<nowiki>作为物理记忆网络和外部源的性质的函数。在上述方程中，α 是“遗忘”时间尺度常数，xi = r-1，r = frac { r _ text { off }{ on }{ r _ text { on }}是记忆电阻器极限电阻的开关和开关值之比，vec s 是电路源的矢量，Omega 是电路基本环路的投影仪。常数 β 具有电压的尺寸，与记忆电阻器的特性有关; 它的物理起源是导体中的电荷迁移率。对角矩阵和向量 w = 操作者名{ diag }(vec w)和 vec w 分别是记忆电阻器的内值，值在0到1之间。因此，这个等式需要在内存值上添加额外的约束，以保证可靠性。</nowiki><br />
<br />
== See also==<br />
== 相关词条==<br />
{{Columns-list|colwidth=18em|<br />
* [[AI accelerator (computer hardware)|AI accelerator]]<br />
* [[Artificial brain]]<br />
* [[Biomorphic]]<br />
* [[Cognitive computer]]<br />
* [[Computation and Neural Systems]]<br />
* [[Differentiable programming]]<br />
* [[Event camera]]<br />
* [[Neurorobotics]]<br />
* [[Optical flow sensor]]<br />
* [[Physical neural network]]<br />
* [[SpiNNaker]]<br />
* [[SyNAPSE]]<br />
* [[Retinomorphic sensor]]<br />
* [[Vision chip]]<br />
* [[Vision processing unit]]<br />
* [[Zeroth (software)]]<br />
* [[Hardware for artificial intelligence]]<br />
}}<br />
<br />
{{Portal bar|Electronics}}<br />
<br />
== References==<br />
{{Reflist|40em}}<br />
<br />
== External links==<br />
<!--======================== {{No more links}} ============================<br />
| PLEASE BE CAUTIOUS IN ADDING MORE LINKS TO THIS ARTICLE. Wikipedia |<br />
| is not a collection of links nor should it be used for advertising. |<br />
| |<br />
| Excessive or inappropriate links WILL BE DELETED. |<br />
| See [[Wikipedia:External links]] & [[Wikipedia:Spam]] for details. |<br />
| |<br />
| If there are already plentiful links, please propose additions or |<br />
| replacements on this article's discussion page, or submit your link |<br />
| to the relevant category at the Open Directory Project (dmoz.org) |<br />
| and link back to that category using the {{dmoz}} template. |<br />
======================={{No more links}}=============================--><br />
{{Commons category|Neuromorphic Engineering}}<br />
*[https://web.archive.org/web/20150727034331/http://ine-web.org/workshops/workshops-overview Telluride Neuromorphic Engineering Workshop]<br />
*[https://archive.today/20130115190057/http://capocaccia.ethz.ch/ CapoCaccia Cognitive Neuromorphic Engineering Workshop]<br />
*[https://web.archive.org/web/20190716132350/http://www.ine-web.org/ Institute of Neuromorphic Engineering]<br />
*[http://www.ine-news.org/ INE news site].<br />
*[http://www.frontiersin.org/neuromorphic_engineering Frontiers in Neuromorphic Engineering Journal]<br />
*[http://www.cns.caltech.edu/ Computation and Neural Systems] department at the [[California Institute of Technology]].<br />
*[http://www.humanbrainproject.eu/ Human Brain Project official site]<br />
*[https://www.the-scientist.com/features/building-a-silicon-brain-65738 Building a Silicon Brain:] Computer chips based on biological neurons may help simulate larger and more-complex brain models. May 1, 2019. SANDEEP RAVINDRAN<br />
<br />
<br />
*Telluride Neuromorphic Engineering Workshop<br />
*CapoCaccia Cognitive Neuromorphic Engineering Workshop<br />
*Institute of Neuromorphic Engineering<br />
*INE news site.<br />
*Frontiers in Neuromorphic Engineering Journal<br />
*Computation and Neural Systems department at the California Institute of Technology.<br />
*Human Brain Project official site<br />
*Building a Silicon Brain: Computer chips based on biological neurons may help simulate larger and more-complex brain models. May 1, 2019. SANDEEP RAVINDRAN<br />
<br />
=<nowiki>外部链接</nowiki>= <br />
*碲化物神经形态工程工作室<br />
*CapoCaccia 认知神经形态工程工作室<br />
* 神经形态工程研究所<br />
*INE 新闻站点。<br />
*《神经形态工程学前沿》<br />
*加州理工学院计算与神经系统系。<br />
*人脑项目官方网站<br />
*打造硅脑: 基于生物神经元的计算机芯片可能有助于模拟更大、更复杂的大脑模型。2019年5月1日。SANDEEP RAVINDRAN<br />
<br />
{{Differentiable computing}}<br />
{{Authority control}}<br />
<br />
[[Category:Electrical engineering]]<br />
[[Category:Neuroscience]]<br />
[[分类:AI accelerators|*]]<br />
[[Category:Artificial intelligence]]<br />
[[Category:Robotics]]<br />
<br />
Category:Electrical engineering<br />
Category:Neuroscience<br />
*<br />
Category:Artificial intelligence<br />
Category:Robotics<br />
<br />
类别: 电气工程类别: 神经科学 * 类别: 人工智能类别: 机器人<br />
<br />
<noinclude><br />
<br />
<small>This page was moved from [[wikipedia:en:Neuromorphic engineering]]. Its edit history can be viewed at [[神经形态计算/edithistory]]</small></noinclude><br />
<br />
[[Category:待整理页面]]<br />
<br />
<references /></div>Shenky20https://wiki.swarma.org/index.php?title=%E7%A5%9E%E7%BB%8F%E5%BD%A2%E6%80%81%E8%AE%A1%E7%AE%97&diff=30184神经形态计算2022-04-14T06:54:06Z<p>Shenky20：/* 案例 */</p>
<hr />
<div><br />
此词条由神经动力学模型读书会词条梳理志愿者Shenky20翻译审校，未经专家审校，带来阅读不便，请见谅。<br />
<br />
{{Use American English|date = January 2019}}<br />
{{Use mdy dates|date = January 2019}}<br />
<br />
'''Neuromorphic engineering''', also known as '''neuromorphic computing''',<ref name=":3">{{Cite journal | doi = 10.1145/2601069| title = Neuromorphic computing gets ready for the (really) big time| journal = [[Communications of the ACM]]| volume = 57| issue = 6| year = 2014| last1 = Monroe | first1 = D. | pages = 13–15| s2cid = 20051102}}</ref><ref name=":4">{{Cite journal | doi = 10.1088/0957-4484/21/17/175202| pmid = 20368686| title = Nanotube devices based crossbar architecture: Toward neuromorphic computing| journal = Nanotechnology| volume = 21| issue = 17| pages = 175202| year = 2010| last1 = Zhao | first1 = W. S. | last2 = Agnus | first2 = G. | last3 = Derycke | first3 = V. | last4 = Filoramo | first4 = A. | last5 = Bourgoin | first5 = J. -P. | last6 = Gamrat | first6 = C. | bibcode = 2010Nanot..21q5202Z| url = https://zenodo.org/record/3428659}}</ref><ref name="humanbrainproject">{{YouTube|id=6RoiZ90mGfw|title=The Human Brain Project SP 9: Neuromorphic Computing Platform}}</ref> is the use of [[very-large-scale integration]] (VLSI) systems containing electronic [[analog circuit]]s to mimic neuro-biological architectures present in the nervous system. A neuromorphic computer/chip is any device that uses physical artificial neurons (made from silicon) to do computations.<ref name=":5">{{cite journal|last1=Mead|first1=Carver|title=Neuromorphic electronic systems|journal=Proceedings of the IEEE|date=1990|volume=78|issue=10|pages=1629–1636|doi=10.1109/5.58356|url=https://authors.library.caltech.edu/53090/1/00058356.pdf}}</ref><ref name=":2" /> In recent times, the term ''neuromorphic'' has been used to describe analog, digital, mixed-mode analog/digital VLSI, and software systems that implement models of [[neural system]]s (for [[perception]], [[motor control]], or [[multisensory integration]]). The implementation of neuromorphic computing on the hardware level can be realized by oxide-based [[memristor]]s,<ref name="Maan 1–13">{{Cite journal|last1=Maan|first1=A. K.|last2=Jayadevi|first2=D. A.|last3=James|first3=A. P.|date=2016-01-01|title=A Survey of Memristive Threshold Logic Circuits|journal=IEEE Transactions on Neural Networks and Learning Systems|volume=PP|issue=99|pages=1734–1746|doi=10.1109/TNNLS.2016.2547842|pmid=27164608|issn=2162-237X|arxiv=1604.07121|bibcode=2016arXiv160407121M|s2cid=1798273}}</ref> spintronic memories, threshold switches, and [[transistor]]s.<ref name=":6">{{Cite journal|title = Mott Memory and Neuromorphic Devices|journal = Proceedings of the IEEE|date = 2015-08-01|issn = 0018-9219|pages = 1289–1310|volume = 103|issue = 8|doi = 10.1109/JPROC.2015.2431914|first1 = You|last1 = Zhou|first2 = S.|last2 = Ramanathan|s2cid = 11347598|url = https://zenodo.org/record/895565}}</ref><ref name=":2">{{Cite document|title=Neuromorphic Circuits With Neural Modulation Enhancing the Information Content of Neural Signaling {{!}} International Conference on Neuromorphic Systems 2020|language=EN|doi=10.1145/3407197.3407204|s2cid=220794387}}</ref> Training software-based neuromorphic systems of spiking neural networks can be achieved using error backpropagation, e.g., using [[Python (programming language)|Python]] based frameworks such as snnTorch,<ref name=":7">{{cite journal |last1=Eshraghian|first1=Jason K.|last2=Ward|first2=Max|last3=Neftci |first3=Emre|last4=Wang|first4=Xinxin|last5=Lenz|first5=Gregor|last6=Dwivedi|first6=Girish|last7=Bennamoun|first7=Mohammed|last8=Jeong|first8=Doo Seok|last9=Lu|first9=Wei D.|title=Training Spiking Neural Networks Using Lessons from Deep Learning |date=1 October 2021 |arxiv=2109.12894 }}</ref> or using canonical learning rules from the biological learning literature, e.g., using BindsNet.<ref name=":8">{{Cite web | url=https://github.com/Hananel-Hazan/bindsnet | title=Hananel-Hazan/bindsnet: Simulation of spiking neural networks (SNNs) using PyTorch.| date=31 March 2020}}</ref><br />
<br />
<br />
'''<font color="#ff8000">神经形态工程Neuromorphic engineering</font>'''(也称为'''<font color="#ff8000">神经形态计算Neuromorphic computing</font>''')<ref name=":3" /><ref name=":4" /><ref name="humanbrainproject" />是指使用包含电子'''<font color="#ff8000">模拟电路Analog circuit</font>'''的'''<font color="#ff8000">超大规模集成电路Very-large-scale integration</font>'''系统来模拟神经系统中生理结构的研究方法。神经形态计算机或神经形态芯片包括任何使用由硅制成的人造神经元进行计算的设备。<ref name=":5" /><ref name=":2" />近年来，神经形态学（neuromorphic）这个术语被用来描述能够实现'''<font color="#ff8000">神经系统Neural system</font>'''模型功能(如'''<font color="#ff8000">感知Perception</font>'''、'''<font color="#ff8000">运动控制Motor control</font>'''，'''<font color="#ff8000">多感官整合Multisensory integration</font>'''等)的模拟、数字、模拟/数字混合模式超大规模集成电路和软件系统。神经形态计算的硬件实现可以通过基于氧化物的'''<font color="#ff8000">记忆电阻器Memristor</font>'''、<ref name="Maan 1–13" />自旋电子存储器、阈值开关和'''<font color="#ff8000">晶体管Transistor</font>'''来实现。<ref name=":6" /><ref name=":2" />对基于软件的脉冲神经网络系统的训练可以通过误差反向传播机制来实现，例如，使用snnTorch等基于Python的框架，<ref name=":7" />或使用BindsNet等典型的受生物启发的学习模式。<ref name=":8" /><br />
<br />
A key aspect of neuromorphic engineering is understanding how the morphology of individual neurons, circuits, applications, and overall architectures creates desirable computations, affects how information is represented, influences robustness to damage, incorporates learning and development, adapts to local change (plasticity), and facilitates evolutionary change.<br />
<br />
<br />
神经形态工程领域的一个关键问题，就是理解'''<font color="#32CD32">单个神经元形态、神经回路、应用和整体结构</font>'''如何产生理想的计算，如何影响信息的表示和对破坏的鲁棒性，如何整合学习和发展，如何适应局部变化(可塑性) 并促进逐渐发展的变化。<br />
<br />
Neuromorphic engineering is an interdisciplinary subject that takes inspiration from [[biology]], [[physics]], [[mathematics]], [[computer science]], and [[electronic engineering]]<ref name=":2" /> to design artificial neural systems, such as [[Machine vision|vision systems]], head-eye systems, auditory processors, and autonomous robots, whose physical architecture and design principles are based on those of biological nervous systems.<ref name=":9">{{Cite journal | doi = 10.1155/2012/705483| title = Qualitative Functional Decomposition Analysis of Evolved Neuromorphic Flight Controllers| journal = Applied Computational Intelligence and Soft Computing| volume = 2012| pages = 1–21| year = 2012| last1 = Boddhu | first1 = S. K. | last2 = Gallagher | first2 = J. C. | doi-access = free}}</ref> It was developed by [[Carver Mead]]<ref name=":10">{{cite web|title=carver mead website|last1=Mead|first1=Carver|url=http://www.carvermead.caltech.edu/index.html|website=carvermead}}</ref> in the late 1980s.<br />
<br />
<br />
神经形态工程是以'''<font color="#ff8000">生物学Biology</font>'''、'''<font color="#ff8000">物理学Physics</font>'''、'''<font color="#ff8000">数学Mathematics</font>'''、'''<font color="#ff8000">计算机科学Computer science</font>'''和'''<font color="#ff8000">电子工程Electronic engineering<ref name=":2" /></font>'''等学科为基础，设计人工神经系统（如'''<font color="#ff8000">视觉系统Vision systems</font>'''、头眼系统、听觉处理器以及物理结构和设计原则都受启发于生物神经系统的自主机器人）的一门交叉学科。<ref name=":9" />20世纪80年代后期，卡弗·米德极大地推动了神经形态工程领域的发展。<ref name=":10" /><br />
<br />
==Neurological inspiration==<br />
==神经生物学启发==<br />
Neuromorphic engineering is set apart by the inspiration it takes from what we know about the structure and operations of the [[brain]]. Neuromorphic engineering translates what we know about the brain's function into computer systems. Work has mostly focused on replicating the analog nature of [[biological computation]] and the role of [[neuron]]s in [[cognition]].<br />
<br />
<br />
神经形态工程形成的灵感来源于现已知的'''<font color="#ff8000">大脑Brain</font>'''结构及运作机制知识，它将我们对大脑功能的了解用于研究和优化计算机系统。该领域工作主要集中于对'''<font color="#ff8000">生物计算Biological computation</font>'''模拟特性和'''<font color="#ff8000">神经元Neuron</font>'''在'''<font color="#ff8000">认知Cognition</font>'''中所发挥作用的复刻。<br />
<br />
The biological processes of neurons and their [[synapse]]s are dauntingly complex, and thus very difficult to artificially simulate. A key feature of biological brains is that all of the processing in neurons use analog chemical signals. This makes it hard to replicate brains in computers because the current generation of computers is completely digital. However, the characteristics of these parts can be abstracted into mathematical functions that closely capture the essence of the neuron's operations.<br />
<br />
神经元及其'''<font color="#ff8000">突触Synapse</font>'''的生理学过程极其复杂，难以进行人工模拟。神经元在所有处理过程中都使用化学模拟信号，这是大脑的一个关键生理学特征。这个特征大大增加了在计算机中复制大脑的难度，因为目前的计算机是完全数字化的。然而，部分特征可以抽象为数学函数，这些函数能够紧密捕捉神经元运行的本质。<br />
<br />
The goal of neuromorphic computing is not to perfectly mimic the brain and all of its functions, but instead to extract what is known of its structure and operations to be used in a practical computing system. No neuromorphic system will claim nor attempt to reproduce every element of neurons and synapses, but all adhere to the idea that computation is highly [[distributed processing|distributed]] throughout a series of small computing elements analogous to a neuron. While this sentiment is standard, researchers chase this goal with different methods.<ref name=":11">{{Cite journal | doi = 10.1088/1741-2560/13/5/051001| title = Large-scale neuromorphic computing systems| journal = Journal of Neural Engineering| volume = 13| pages = 1–15| year = 2016| last1 = Furber | first1 = Steve| issue = 5| pmid = 27529195| bibcode = 2016JNEng..13e1001F| doi-access = free}}</ref><br />
<br />
神经形态计算的目标不是完美地模拟大脑及其所有功能，而是利用已知的大脑结构和运转机制来研发或优化实际的计算系统。任何神经形态学系统都不会声称或试图复制神经元和突触中的每一个元素，但所有人都一致认可将计算高度'''<font color="#ff8000">分散Distribute</font>'''于一系列类似于神经元的小型计算元素的理念。研究人员用不同的方法来追求这一普遍目标。<ref name=":11" /><br />
<br />
<br />
==Examples==<br />
==代表性成果==<br />
As early as 2006, researchers at [[Georgia Tech]] published a field programmable neural array.<ref name=":12">{{Cite book|title = A field programmable neural array|last1 = Farquhar|first1 = Ethan|date = May 2006|journal = IEEE International Symposium on Circuits and Systems|pages = 4114–4117|last2 = Hasler|first2 = Paul.|doi = 10.1109/ISCAS.2006.1693534|isbn = 978-0-7803-9389-9|s2cid = 206966013}}</ref> This chip was the first in a line of increasingly complex arrays of floating gate transistors that allowed programmability of charge on the gates of [[MOSFET]]s to model the channel-ion characteristics of neurons in the brain and was one of the first cases of a silicon programmable array of neurons.<br />
<br />
早在2006年，佐治亚理工学院的研究人员就研发出了一种现场可编程神经阵列。<ref name=":12" />在此之后，一系列越来越复杂的浮栅晶体管阵列被成功研发出来，这些晶体管阵列可以通过在'''<font color="#ff8000">金属-氧化物半导体效应晶体管MOSFET</font>'''的栅极上编程来模拟大脑中神经元的离子通道特性，这也是以硅为主要材料的可编程神经元阵列的最早成功案例之一。<br />
<br />
In November 2011, a group of [[MIT]] researchers created a computer chip that mimics the analog, ion-based communication in a synapse between two neurons using 400 transistors and standard [[CMOS]] manufacturing techniques.<ref name=":13">{{cite web|title=MIT creates "brain chip"|url=http://www.extremetech.com/extreme/105067-mit-creates-brain-chip|access-date=4 December 2012}}</ref><ref name="Neuromorphic silicon paper">{{cite journal|title=Neuromorphic silicon neurons and large-scale neural networks: challenges and opportunities|doi=10.3389/fnins.2011.00108|pmid=21991244|pmc=3181466|volume=5|pages=108|journal=Frontiers in Neuroscience|year=2011|last1=Poon|first1=Chi-Sang|last2=Zhou|first2=Kuan|doi-access=free}}</ref><br />
<br />
2011年11月，麻省理工学院的一组研究人员研发出一种计算机芯片，该芯片上使用标准的'''<font color="#ff8000">互补金属氧化物半导体CMOS</font>'''制造技术集成了400个晶体管，能够模拟神经元间突触中基于离子的通讯。<ref name=":13" /><ref name="Neuromorphic silicon paper" /><br />
<br />
In June 2012, [[spintronic]] researchers at [[Purdue University]] presented a paper on the design of a neuromorphic chip using [[Spin valve|lateral spin valve]]s and [[memristor]]s. They argue that the architecture works similarly to neurons and can therefore be used to test methods of reproducing the brain's processing. In addition, these chips are significantly more energy-efficient than conventional ones.<ref name="Spin Devices Prop">{{Cite arXiv|title=Proposal For Neuromorphic Hardware Using Spin Devices|eprint=1206.3227|last1=Sharad|first1=Mrigank|last2=Augustine|first2=Charles|last3=Panagopoulos|first3=Georgios|last4=Roy|first4=Kaushik|class=cond-mat.dis-nn|year=2012}}</ref><br />
<br />
2012年6月，普渡大学的'''<font color="#ff8000">自旋电子学Spintronic</font>'''研究人员发表了一篇关于利用'''<font color="#ff8000">侧向自旋阀Lateral spin valves</font>'''和'''<font color="#ff8000">记忆电阻器Memristors</font>'''设计神经形态芯片的论文。他们认为，这种芯片结构的工作原理与神经元相似，因此可以用于大脑运行机制的复刻方法的测试。此外，这些芯片在能耗方面明显优于传统芯片。<ref name="Spin Devices Prop" /><br />
<br />
Research at [[HP Labs]] on Mott memristors has shown that while they can be non-[[Volatile memory|volatile]], the volatile behavior exhibited at temperatures significantly below the [[phase transition]] temperature can be exploited to fabricate a [[neuristor]],<ref name=":0" /> a biologically-inspired device that mimics behavior found in neurons.<ref name=":0">{{Cite journal | doi = 10.1038/nmat3510| pmid = 23241533| title = A scalable neuristor built with Mott memristors| journal = Nature Materials| volume = 12| issue = 2| pages = 114–7| year = 2012| last1 = Pickett | first1 = M. D. | last2 = Medeiros-Ribeiro | first2 = G. | last3 = Williams | first3 = R. S. | bibcode = 2013NatMa..12..114P| s2cid = 16271627| url = https://semanticscholar.org/paper/b6ba6f496ace2b947f111059663e76bb60e9efeb}}</ref> In September 2013, they presented models and simulations that show how the spiking behavior of these neuristors can be used to form the components required for a [[Turing machine]].<ref name=":14">{{cite journal|doi=10.1088/0957-4484/24/38/384002|title=Phase transitions enable computational universality in neuristor-based cellular automata|author1=Matthew D Pickett|author2=R Stanley Williams|name-list-style=amp|date=September 2013|publisher=IOP Publishing Ltd|journal=Nanotechnology|volume=24|issue=38|pmid=23999059|bibcode=2013Nanot..24L4002P|at=384002}}</ref><br />
<br />
'''<font color="#ff8000">惠普实验室HP labs</font>'''在莫特记忆电阻器上的研究表明，尽管它们可以是非'''<font color="#ff8000">易失性Volatile</font>'''的，但是在'''<font color="#ff8000">相变Phase transition</font>'''温度以下时表现出的易失性行为可以被用来制造'''<font color="#ff8000">类神经元电阻器Neuristor</font>'''(一种生物学启发的模仿神经元行为的硬件)<ref name=":0" />。2013年9月，他们通过模型和仿真展示了这些类神经元电阻器的脉冲行为如何产生'''<font color="#ff8000">图灵机Turing machine</font>'''的所需元素。<ref name=":14" /><br />
<br />
[[Neurogrid]], built by ''Brains in Silicon'' at [[Stanford University]],<ref name=":15">{{cite journal|last1=Boahen|first1=Kwabena|title=Neurogrid: A Mixed-Analog-Digital Multichip System for Large-Scale Neural Simulations|journal=Proceedings of the IEEE|date=24 April 2014|volume=102|issue=5|pages=699–716|doi=10.1109/JPROC.2014.2313565|s2cid=17176371}}</ref> is an example of hardware designed using neuromorphic engineering principles. The circuit board is composed of 16 custom-designed chips, referred to as NeuroCores. Each NeuroCore's analog circuitry is designed to emulate neural elements for 65536 neurons, maximizing energy efficiency. The emulated neurons are connected using digital circuitry designed to maximize spiking throughput.<ref name=":16">{{cite journal|doi=10.1038/503022a|pmid = 24201264|title = Neuroelectronics: Smart connections|journal = Nature|volume = 503|issue = 7474|pages = 22–4|year = 2013|last1 = Waldrop|first1 = M. Mitchell|bibcode = 2013Natur.503...22W|doi-access = free}}</ref><ref name=":17">{{cite journal|doi=10.1109/JPROC.2014.2313565|title = Neurogrid: A Mixed-Analog-Digital Multichip System for Large-Scale Neural Simulations|journal = Proceedings of the IEEE|volume = 102|issue = 5|pages = 699–716|year = 2014|last1 = Benjamin|first1 = Ben Varkey|last2 = Peiran Gao|last3 = McQuinn|first3 = Emmett|last4 = Choudhary|first4 = Swadesh|last5 = Chandrasekaran|first5 = Anand R.|last6 = Bussat|first6 = Jean-Marie|last7 = Alvarez-Icaza|first7 = Rodrigo|last8 = Arthur|first8 = John V.|last9 = Merolla|first9 = Paul A.|last10 = Boahen|first10 = Kwabena|s2cid = 17176371}}</ref><br />
<br />
'''<font color="#ff8000">神经栅格Neurogrid<ref name=":15" /></font>'''是由斯坦福大学Brains in Silicon公司研发的、使用神经形态工程原理设计的硬件。该电路板由16个定制设计的芯片组成(NeuroCores)。在设计中，每个NeuroCore芯片的模拟电路对65536个神经元的神经元素进行模拟，以最大限度地提高能量效率。模拟出的神经元通过设计的数字电路连接，以最大化脉冲吞吐量。<ref name=":16" /><ref name=":17" /><br />
<br />
A research project with implications for neuromorphic engineering is the [[Human Brain Project]] that is attempting to simulate a complete human brain in a supercomputer using biological data. It is made up of a group of researchers in neuroscience, medicine, and computing.<ref>{{cite web|title=Involved Organizations|url=http://www.humanbrainproject.eu/partners.html|access-date=22 February 2013|url-status=dead|archive-url=https://web.archive.org/web/20130302142627/http://www.humanbrainproject.eu/partners.html|archive-date=2 March 2013}}</ref> [[Henry Markram]], the project's co-director, has stated that the project proposes to establish a foundation to explore and understand the brain and its diseases, and to use that knowledge to build new computing technologies. The three primary goals of the project are to better understand how the pieces of the brain fit and work together, to understand how to objectively diagnose and treat brain diseases, and to use the understanding of the human brain to develop neuromorphic computers. That the simulation of a complete human brain will require a supercomputer a thousand times more powerful than today's encourages the current focus on neuromorphic computers.<ref>{{cite web|title=Human Brain Project|url=http://www.humanbrainproject.eu|access-date=22 February 2013}}</ref> $1.3 billion has been allocated to the project by The [[European Commission]].<ref>{{cite web|title=The Human Brain Project and Recruiting More Cyberwarriors|url=http://www.marketplace.org/topics/tech/human-brain-project-and-recruiting-more-cyberwarriors|access-date=22 February 2013|date=January 29, 2013}}</ref><br />
<br />
A research project with implications for neuromorphic engineering is the Human Brain Project that is attempting to simulate a complete human brain in a supercomputer using biological data. It is made up of a group of researchers in neuroscience, medicine, and computing. Henry Markram, the project's co-director, has stated that the project proposes to establish a foundation to explore and understand the brain and its diseases, and to use that knowledge to build new computing technologies. The three primary goals of the project are to better understand how the pieces of the brain fit and work together, to understand how to objectively diagnose and treat brain diseases, and to use the understanding of the human brain to develop neuromorphic computers. That the simulation of a complete human brain will require a supercomputer a thousand times more powerful than today's encourages the current focus on neuromorphic computers. $1.3 billion has been allocated to the project by The European Commission.<br />
<br />
一个对神经形态工程有影响的研究项目是人脑项目，它试图用生物数据在超级计算机中模拟完整的人脑。它由一群神经科学、医学和计算机科学的研究人员组成。该项目的联合主管亨利•马克拉姆(Henry Markram)表示，该项目提议建立一个基础，以探索和了解大脑及其疾病，并利用这些知识构建新的计算机技术。这个项目的三个主要目标是: 更好地理解大脑的各个部分是如何相互匹配和协同工作的; 理解如何客观地诊断和治疗脑部疾病; 以及利用对人类大脑的理解来开发神经形态计算机。模拟一个完整的人类大脑需要一台比现在强大一千倍的超级计算机，这鼓励了当前对神经形态计算机的关注。欧洲委员会已经拨款13亿美元用于这个项目。<br />
<br />
Other research with implications for neuromorphic engineering involves the [[BRAIN Initiative]]<ref name="economist">[https://www.economist.com/news/science-and-technology/21582495-computers-will-help-people-understand-brains-better-and-understanding-brains Neuromorphic computing: The machine of a new soul], The Economist, 2013-08-03</ref> and the [[TrueNorth]] chip from [[IBM]].<ref>{{cite journal|last1=Modha|first1=Dharmendra|title=A million spiking-neuron integrated circuit with a scalable communication network and interface|journal=Science|date=Aug 2014|volume=345|issue=6197|pages=668–673|doi=10.1126/science.1254642|pmid=25104385|bibcode=2014Sci...345..668M|s2cid=12706847}}</ref> Neuromorphic devices have also been demonstrated using nanocrystals, nanowires, and conducting polymers.<ref>{{Cite web|url=http://jessamynfairfield.com/wp-content/uploads/2017/03/PWMar17Fairfield.pdf|title=Smarter Machines|last=Fairfield|first=Jessamyn|date=March 1, 2017}}</ref><br />
<br />
Other research with implications for neuromorphic engineering involves the BRAIN InitiativeNeuromorphic computing: The machine of a new soul, The Economist, 2013-08-03 and the TrueNorth chip from IBM. Neuromorphic devices have also been demonstrated using nanocrystals, nanowires, and conducting polymers.<br />
<br />
其他与神经形态工程有关的研究还包括 BRAIN initiativity/euromorphic computing: The machine of a new soul，The Economist，2013-08-03 and The TrueNorth chip from IBM。神经形态学设备也已经被证明使用纳米晶体、纳米线和导电聚合物。<br />
<br />
[[Intel]] unveiled its neuromorphic research chip, called “[[Intel Loihi|Loihi]]”, in October 2017. The chip uses an asynchronous [[spiking neural network]] (SNN) to implement adaptive self-modifying event-driven fine-grained parallel computations used to implement learning and inference with high efficiency.<ref>{{cite journal |last1=Davies |first1=Mike |title=Loihi: A Neuromorphic Manycore Processor with On-Chip Learning |journal=IEEE Micro |date=January 16, 2018 |volume=38 |issue=1 |pages=82–99 |display-authors=etal|doi=10.1109/MM.2018.112130359 |s2cid=3608458 }}</ref><ref name="Morris2017">{{cite web |last1=Morris |first1=John |title=Why Intel built a neuromorphic chip |url=https://www.zdnet.com/article/why-intel-built-a-neuromorphic-chip/ |website=ZDNet |access-date=17 August 2018 |language=en}}</ref><br />
<br />
Intel unveiled its neuromorphic research chip, called “Loihi”, in October 2017. The chip uses an asynchronous spiking neural network (SNN) to implement adaptive self-modifying event-driven fine-grained parallel computations used to implement learning and inference with high efficiency.<br />
<br />
2017年10月，英特尔公布了其神经形态研究芯片“ Loihi”。该芯片采用异步脉冲神经网络(SNN)实现自适应自修改事件驱动的细粒度并行计算，实现了高效的学习和推理。<br />
<br />
[[IMEC]], a Belgium-based nanoelectronics research center, demonstrated the world's first self-learning neuromorphic chip. The brain-inspired chip, based on OxRAM technology, has the capability of self-learning and has been demonstrated to have the ability to compose music.<ref>{{cite web |title=Imec demonstrates self-learning neuromorphic chip that composes music |url=https://www.imec-int.com/en/articles/imec-demonstrates-self-learning-neuromorphic-chip-that-composes-music |website=IMEC International |access-date=1 October 2019}}</ref> IMEC released the 30-second tune composed by the prototype. The chip was sequentially loaded with songs in the same time signature and style. The songs were old Belgian and French flute minuets, from which the chip learned the rules at play and then applied them.<ref>{{cite web|last1=Bourzac|first1=Katherine|title=A Neuromorphic Chip That Makes Music|url=https://spectrum.ieee.org/a-neuromorphic-chip-that-makes-music|url-status=live|access-date=1 October 2019|website=IEEE Spectrum|date=May 23, 2017}}</ref><br />
<br />
IMEC, a Belgium-based nanoelectronics research center, demonstrated the world's first self-learning neuromorphic chip. The brain-inspired chip, based on OxRAM technology, has the capability of self-learning and has been demonstrated to have the ability to compose music. IMEC released the 30-second tune composed by the prototype. The chip was sequentially loaded with songs in the same time signature and style. The songs were old Belgian and French flute minuets, from which the chip learned the rules at play and then applied them.<br />
<br />
比利时的一个纳米电子学研究中心 IMEC 展示了世界上第一个自学习神经形态芯片。这种基于 OxRAM 技术的大脑启发芯片具有自学习能力，并已被证明具有创作音乐的能力。IMEC 发布了由原型机谱写的30秒曲调。芯片按顺序加载同时签名和风格的歌曲。歌曲是古老的比利时和法国长笛小步舞曲，筹码从中学习游戏规则，然后应用它们。<br />
<br />
[[Blue Brain Project|The Blue Brain Project]], led by Henry Markram, aims to build biologically detailed digital reconstructions and simulations of the mouse brain. The Blue Brain Project has created in silico models of rodent brains, while attempting to replicate as many details about its biology as possible. The supercomputer-based simulations offer new perspectives on understanding the structure and functions of the brain.<br />
<br />
The Blue Brain Project, led by Henry Markram, aims to build biologically detailed digital reconstructions and simulations of the mouse brain. The Blue Brain Project has created in silico models of rodent brains, while attempting to replicate as many details about its biology as possible. The supercomputer-based simulations offer new perspectives on understanding the structure and functions of the brain.<br />
<br />
由亨利 · 马克拉姆领导的蓝脑计划旨在建立小鼠大脑的生物细节数字重建和模拟。蓝色大脑项目已经在啮齿动物大脑的电子模型中创建，同时试图尽可能多地复制其生物学细节。基于超级计算机的模拟为理解大脑的结构和功能提供了新的视角。<br />
<br />
The European Union funded a series of projects at the University of Heidelberg, which led to the development of <br />
[[BrainScaleS]] (brain-inspired multiscale computation in neuromorphic hybrid systems), a hybrid analog [[neuromorphic]] supercomputer located at Heidelberg University, Germany. It was developed as part of the [[Human Brain Project]] neuromorphic computing platform and is the complement to the [[SpiNNaker]] supercomputer (which is based on digital technology). The architecture used in BrainScaleS mimics biological neurons and their connections on a physical level; additionally, since the components are made of silicon, these model neurons operate on average 864 times (24 hours of real time is 100 seconds in the machine simulation) that of their biological counterparts.<ref>{{Cite web|date=2016-03-21|title=Beyond von Neumann, Neuromorphic Computing Steadily Advances|url=https://www.hpcwire.com/2016/03/21/lacking-breakthrough-neuromorphic-computing-steadily-advance/|access-date=2021-10-08|website=HPCwire|language=en-US}}</ref><br />
<br />
The European Union funded a series of projects at the University of Heidelberg, which led to the development of <br />
BrainScaleS (brain-inspired multiscale computation in neuromorphic hybrid systems), a hybrid analog neuromorphic supercomputer located at Heidelberg University, Germany. It was developed as part of the Human Brain Project neuromorphic computing platform and is the complement to the SpiNNaker supercomputer (which is based on digital technology). The architecture used in BrainScaleS mimics biological neurons and their connections on a physical level; additionally, since the components are made of silicon, these model neurons operate on average 864 times (24 hours of real time is 100 seconds in the machine simulation) that of their biological counterparts.<br />
<br />
欧盟资助了海德堡大学的一系列项目，这些项目导致了 BrainScaleS (神经形态混合系统中受大脑启发的多尺度计算)的发展，这是一台位于德国海德堡大学的混合模拟神经形态超级计算机。它是作为人脑计划神经形态计算平台的一部分而开发的，是 SpiNNaker 超级计算机(基于数字技术)的补充。大脑尺度中使用的体系结构模拟了生物神经元及其在物理层面上的连接; 此外，由于这些组件是由硅制成的，这些模型神经元平均运行864倍(在机器模拟中，24小时的实时时间是100秒)。<br />
<br />
===Neuromorphic sensors ===<br />
===神经形态传感器===<br />
The concept of neuromorphic systems can be extended to sensors (not just to computation). An example of this applied to detecting [[light]] is the [[retinomorphic sensor]] or, when employed in an array, the [[event camera]].<br />
<br />
The concept of neuromorphic systems can be extended to sensors (not just to computation). An example of this applied to detecting light is the retinomorphic sensor or, when employed in an array, the event camera.<br />
<br />
神经形态系统的概念可以扩展到传感器(而不仅仅是计算)。用于检测光线的一个例子是视网膜变形传感器，或者在阵列中使用的事件摄像机。<br />
<br />
<br />
==Ethical considerations==<br />
==伦理问题==<br />
While the interdisciplinary concept of neuromorphic engineering is relatively new, many of the same ethical considerations apply to neuromorphic systems as apply to [[human-like machines]] and [[artificial intelligence]] in general. However, the fact that neuromorphic systems are designed to mimic a [[human brain]] gives rise to unique ethical questions surrounding their usage.<br />
<br />
While the interdisciplinary concept of neuromorphic engineering is relatively new, many of the same ethical considerations apply to neuromorphic systems as apply to human-like machines and artificial intelligence in general. However, the fact that neuromorphic systems are designed to mimic a human brain gives rise to unique ethical questions surrounding their usage.<br />
<br />
虽然神经形态工程的跨学科概念相对较新，但许多同样的伦理考虑适用于神经形态系统，就像适用于类人机器和一般人工智能一样。然而，神经形态系统是为了模仿人类大脑而设计的这一事实引起了围绕其使用的独特的伦理问题。<br />
<br />
However, the practical debate is that neuromorphic hardware as well as artificial "neural networks" are immensely simplified models of how the brain operates or processes information at a much lower [[complex system|complexity in terms of size and functional technology]] and a much more regular structure in terms of [[brain connectivity|connectivity]]. Comparing [[neuromorphic chip]]s to the brain is a very crude comparison similar to comparing a plane to a bird just because they both have wings and a tail. The fact is that neural cognitive systems are many orders of magnitude more [[Energy efficiency (physics)|energy-]] and compute-efficient than current state-of-the-art AI and neuromorphic engineering is an attempt to narrow this gap by inspiring from the brain's mechanism just like many engineering designs have [[bioengineering|bio-inspired features]].<br />
<br />
However, the practical debate is that neuromorphic hardware as well as artificial "neural networks" are immensely simplified models of how the brain operates or processes information at a much lower complexity in terms of size and functional technology and a much more regular structure in terms of connectivity. Comparing neuromorphic chips to the brain is a very crude comparison similar to comparing a plane to a bird just because they both have wings and a tail. The fact is that neural cognitive systems are many orders of magnitude more energy- and compute-efficient than current state-of-the-art AI and neuromorphic engineering is an attempt to narrow this gap by inspiring from the brain's mechanism just like many engineering designs have bio-inspired features.<br />
<br />
然而，实际上的争论是，神经形态硬件和人工“神经网络”是大脑如何运作或处理信息的极其简化的模型，在大小和功能技术方面的复杂性要低得多，在连接方面的结构则更加规则。将神经形态芯片与大脑进行比较是一种非常粗糙的比较，类似于仅仅因为一架飞机有翅膀和一条尾巴就将它与一只鸟进行比较。事实上，神经认知系统比当前最先进的人工智能具有更多的能量和计算效率，而神经形态工程是一种通过从大脑机制中激发灵感来缩小这种差距的尝试，就像许多工程设计具有生物启发的特征一样。数量级。<br />
<br />
===Democratic concerns ===<br />
===公众担忧===<br />
Significant ethical limitations may be placed on neuromorphic engineering due to public perception.<ref>{{Cite report|url=https://ai100.stanford.edu/sites/g/files/sbiybj9861/f/ai_100_report_0831fnl.pdf|title=Artificial Intelligence and Life in 2030|author=2015 Study Panel|date=September 2016|work=One Hundred Year Study on Artificial Intelligence (AI100)|publisher=Stanford University}}</ref> Special [[Eurobarometer]] 382: Public Attitudes Towards Robots, a survey conducted by the European Commission, found that 60% of [[European Union]] citizens wanted a ban of robots in the care of children, the elderly, or the disabled. Furthermore, 34% were in favor of a ban on robots in education, 27% in healthcare, and 20% in leisure. The European Commission classifies these areas as notably “human.” The report cites increased public concern with robots that are able to mimic or replicate human functions. Neuromorphic engineering, by definition, is designed to replicate the function of the human brain.<ref name=":1">{{Cite web|url=http://ec.europa.eu/commfrontoffice/publicopinion/archives/ebs/ebs_382_en.pdf|title=Special Eurobarometer 382: Public Attitudes Towards Robots|last=European Commission|date=September 2012|website=European Commission}}</ref><br />
<br />
Significant ethical limitations may be placed on neuromorphic engineering due to public perception. Special Eurobarometer 382: Public Attitudes Towards Robots, a survey conducted by the European Commission, found that 60% of European Union citizens wanted a ban of robots in the care of children, the elderly, or the disabled. Furthermore, 34% were in favor of a ban on robots in education, 27% in healthcare, and 20% in leisure. The European Commission classifies these areas as notably “human.” The report cites increased public concern with robots that are able to mimic or replicate human functions. Neuromorphic engineering, by definition, is designed to replicate the function of the human brain.<br />
<br />
由于公众的认知，神经形态工程学可能会受到严重的伦理限制。欧盟委员会进行的一项调查发现，60% 的欧盟公民希望禁止照顾儿童、老人或残疾人的机器人。此外，34% 的人支持禁止机器人用于教育，27% 的人支持医疗保健，20% 的人支持休闲。欧盟委员会将这些地区明显归类为“人类”报告指出，公众越来越关注能够模仿或复制人类功能的机器人。神经形态工程，顾名思义，是为了复制人脑的功能而设计的。<br />
<br />
The democratic concerns surrounding neuromorphic engineering are likely to become even more profound in the future. The European Commission found that EU citizens between the ages of 15 and 24 are more likely to think of robots as human-like (as opposed to instrument-like) than EU citizens over the age of 55. When presented an image of a robot that had been defined as human-like, 75% of EU citizens aged 15–24 said it corresponded with the idea they had of robots while only 57% of EU citizens over the age of 55 responded the same way. The human-like nature of neuromorphic systems, therefore, could place them in the categories of robots many EU citizens would like to see banned in the future.<ref name=":1" /><br />
<br />
The democratic concerns surrounding neuromorphic engineering are likely to become even more profound in the future. The European Commission found that EU citizens between the ages of 15 and 24 are more likely to think of robots as human-like (as opposed to instrument-like) than EU citizens over the age of 55. When presented an image of a robot that had been defined as human-like, 75% of EU citizens aged 15–24 said it corresponded with the idea they had of robots while only 57% of EU citizens over the age of 55 responded the same way. The human-like nature of neuromorphic systems, therefore, could place them in the categories of robots many EU citizens would like to see banned in the future.<br />
<br />
围绕神经形态工程的民主关注可能在未来变得更加深刻。欧盟委员会(European Commission)发现，15至24岁的欧盟公民比55岁以上的欧盟公民更有可能认为机器人像人(而不是像仪器)。当看到一张被定义为“类人”的机器人图片时，年龄在15岁至24岁之间的欧盟公民中有75% 的人表示，这与他们对机器人的想法相符，而55岁以上的欧盟公民中只有57% 的人有同样的反应。因此，类似人类的神经形态系统，可以把它们归入许多欧盟公民希望在未来禁止使用的机器人类别。<br />
<br />
===Personhood===<br />
=== 人格问题===<br />
As neuromorphic systems have become increasingly advanced, some scholars{{who|date=August 2021}} have advocated for granting [[personhood]] rights to these systems. If the brain is what grants humans their personhood, to what extent does a neuromorphic system have to mimic the human brain to be granted personhood rights? Critics of technology development in the [[Human Brain Project]], which aims to advance brain-inspired computing, have argued that advancement in neuromorphic computing could lead to machine consciousness or personhood.<ref>{{Cite journal|last=Aicardi|first=Christine|date=September 2018|title=Accompanying technology development in the Human Brain Project: From foresight to ethics management|journal=Futures|volume=102|pages=114–124|doi=10.1016/j.futures.2018.01.005|doi-access=free}}</ref> If these systems are to be treated as people, critics argue, then many tasks humans perform using neuromorphic systems, including the act of termination of neuromorphic systems, may be morally impermissible as these acts would violate the autonomy of the neuromorphic systems.<ref>{{Cite journal|last=Lim|first=Daniel|date=2014-06-01|title=Brain simulation and personhood: a concern with the Human Brain Project|journal=Ethics and Information Technology|language=en|volume=16|issue=2|pages=77–89|doi=10.1007/s10676-013-9330-5|s2cid=17415814|issn=1572-8439}}</ref><br />
<br />
As neuromorphic systems have become increasingly advanced, some scholars have advocated for granting personhood rights to these systems. If the brain is what grants humans their personhood, to what extent does a neuromorphic system have to mimic the human brain to be granted personhood rights? Critics of technology development in the Human Brain Project, which aims to advance brain-inspired computing, have argued that advancement in neuromorphic computing could lead to machine consciousness or personhood. If these systems are to be treated as people, critics argue, then many tasks humans perform using neuromorphic systems, including the act of termination of neuromorphic systems, may be morally impermissible as these acts would violate the autonomy of the neuromorphic systems.<br />
<br />
随着神经形态系统的日益发展，一些学者主张赋予这些系统人格权。如果是大脑赋予了人类人格，那么神经形态系统在多大程度上必须模仿人类大脑才能被赋予人格权利？“人脑计划”旨在推进以大脑为灵感的计算机技术发展，该计划的批评者认为，神经形态计算机技术的进步可能导致机器意识或人格的形成。批评家认为，如果这些系统被当作人来对待，那么人类使用神经形态系统执行的许多任务，包括终止神经形态系统的行为，可能在道德上是不允许的，因为这些行为将违反神经形态系统的自主性。<br />
<br />
==Dual use (military applications)==<br />
==军民两用技术==<br />
The [[Joint Artificial Intelligence Center]], a branch of the U.S. military, is a center dedicated to the procurement and implementation of AI software and neuromorphic hardware for combat use. Specific applications include smart headsets/goggles and robots. JAIC intends to rely heavily on neuromorphic technology to connect "every fighter every shooter" within a network of neuromorphic-enabled units.<br />
<br />
The Joint Artificial Intelligence Center, a branch of the U.S. military, is a center dedicated to the procurement and implementation of AI software and neuromorphic hardware for combat use. Specific applications include smart headsets/goggles and robots. JAIC intends to rely heavily on neuromorphic technology to connect "every fighter every shooter" within a network of neuromorphic-enabled units.<br />
<br />
军民两用联合人工智能中心是美国军队的一个分支，专门从事采购和实施用于战斗的人工智能软件和神经形态硬件。具体应用包括智能耳机/护目镜和机器人。JAIC 打算严重依赖神经形态技术来连接神经形态单位网络中的“每个战士每个射手”。<br />
<br />
==Legal considerations==<br />
==法律问题==<br />
Skeptics have argued that there is no way to apply the electronic personhood, the concept of personhood that would apply to neuromorphic technology, legally. In a letter signed by 285 experts in law, robotics, medicine, and ethics opposing a European Commission proposal to recognize “smart robots” as legal persons, the authors write, “A legal status for a robot can’t derive from the [[Natural person|Natural Person]] model, since the robot would then hold [[human rights]], such as the right to dignity, the right to its integrity, the right to remuneration or the right to citizenship, thus directly confronting the Human rights. This would be in contradiction with the [[Charter of Fundamental Rights of the European Union]] and the [[Convention for the Protection of Human Rights and Fundamental Freedoms]].”<ref>{{Cite web|url=http://www.robotics-openletter.eu/|title=Robotics Openletter {{!}} Open letter to the European Commission|language=fr-FR|access-date=2019-05-10}}</ref><br />
<br />
Skeptics have argued that there is no way to apply the electronic personhood, the concept of personhood that would apply to neuromorphic technology, legally. In a letter signed by 285 experts in law, robotics, medicine, and ethics opposing a European Commission proposal to recognize “smart robots” as legal persons, the authors write, “A legal status for a robot can’t derive from the Natural Person model, since the robot would then hold human rights, such as the right to dignity, the right to its integrity, the right to remuneration or the right to citizenship, thus directly confronting the Human rights. This would be in contradiction with the Charter of Fundamental Rights of the European Union and the Convention for the Protection of Human Rights and Fundamental Freedoms.”<br />
<br />
怀疑论者认为，在法律上没有办法应用电子人格，这个人格概念将适用于神经形态技术。在一封由285名法律、机器人、医学和伦理学专家签名的信中，作者们反对欧盟委员会承认“智能机器人”为法人的提议，他们写道，“机器人的法律地位不能从自然人模型中推导出来，因为机器人将拥有人权，如尊严权、完整权、报酬权或公民权，从而直接面对人权。这将有悖于《欧洲联盟基本权利宪章和《保护人权和基本自由公约》。”<br />
<br />
===Ownership and property rights===<br />
===所有权及财产权问题===<br />
There is significant legal debate around property rights and artificial intelligence. In ''Acohs Pty Ltd v. Ucorp Pty Ltd'', Justice Christopher Jessup of the [[Federal Court of Australia]] found that the [[source code]] for [[Material safety data sheets|Material Safety Data Sheets]] could not be [[Copyright law of Australia|copyrighted]] as it was generated by a [[software interface]] rather than a human author.<ref>{{Cite web|url=http://www.lavan.com.au/advice/intellectual_property_technology/copyright_in_source_code_and_digital_products|title=Copyright in source code and digital products|last=Lavan|website=Lavan|language=en|access-date=2019-05-10}}</ref> The same question may apply to neuromorphic systems: if a neuromorphic system successfully mimics a human brain and produces a piece of original work, who, if anyone, should be able to claim ownership of the work?<ref>{{cite journal |last1=Eshraghian|first1=Jason K. |title=Human Ownership of Artificial Creativity |journal=Nature Machine Intelligence |date=9 March 2020 |volume=2 |pages=157–160 |doi=10.1038/s42256-020-0161-x}}</ref><br />
<br />
There is significant legal debate around property rights and artificial intelligence. In Acohs Pty Ltd v. Ucorp Pty Ltd, Justice Christopher Jessup of the Federal Court of Australia found that the source code for Material Safety Data Sheets could not be copyrighted as it was generated by a software interface rather than a human author. The same question may apply to neuromorphic systems: if a neuromorphic system successfully mimics a human brain and produces a piece of original work, who, if anyone, should be able to claim ownership of the work?<br />
<br />
围绕财产权和人工智能有着重大的法律争论。在 Acohs Pty Ltd 诉 Ucorp Pty Ltd 一案中，澳大利亚联邦法院的克里斯托弗 · 杰瑟普法官发现，材料安全数据表的源代码不能受版权保护，因为它是由软件界面而不是人工作者生成的。同样的问题可能也适用于神经形态系统: 如果一个神经形态系统成功地模仿了人类的大脑并产生了一部原创作品，那么谁，如果有人，应该声称拥有这部作品的所有权？<br />
<br />
==Neuromemristive systems==<br />
==神经记忆电阻系统==<br />
Neuromemristive systems are a subclass of neuromorphic computing systems that focus on the use of [[memristors]] to implement [[neuroplasticity]]. While neuromorphic engineering focuses on mimicking biological behavior, neuromemristive systems focus on abstraction.<ref>{{Cite web|url=https://digitalops.sandia.gov/Mediasite/Play/a10cf6ceb55d47608bb8326dd00e46611d|title=002.08 N.I.C.E. Workshop 2014: Towards Intelligent Computing with Neuromemristive Circuits and Systems - Feb. 2014|website=digitalops.sandia.gov|access-date=2019-08-26}}</ref> For example, a neuromemristive system may replace the details of a [[Cerebral cortex|cortical]] microcircuit's behavior with an abstract neural network model.<ref>C. Merkel and D. Kudithipudi, "Neuromemristive extreme learning machines for pattern classification," ISVLSI, 2014.</ref><br />
<br />
Neuromemristive systems are a subclass of neuromorphic computing systems that focus on the use of memristors to implement neuroplasticity. While neuromorphic engineering focuses on mimicking biological behavior, neuromemristive systems focus on abstraction. For example, a neuromemristive system may replace the details of a cortical microcircuit's behavior with an abstract neural network model.C. Merkel and D. Kudithipudi, "Neuromemristive extreme learning machines for pattern classification," ISVLSI, 2014.<br />
<br />
神经记忆电阻系统是神经形态计算系统的一个亚类，主要研究利用记忆电阻器实现神经可塑性。神经形态工程的重点是模拟生物行为，而神经记忆电阻系统的重点是提取。例如，一个神经记忆系统可以用一个抽象的神经网络模型取代皮层微电路的行为细节。默克尔和 d. Kudithipudi，“用于模式分类的神经记忆极端学习机器”，ISVLSI，2014。<br />
<br />
There exist several neuron inspired threshold logic functions<ref name="Maan 1–13" /> implemented with memristors that have applications in high level [[pattern recognition]] applications. Some of the applications reported recently include [[speech recognition]],<ref>{{Cite journal|title = Memristor pattern recogniser: isolated speech word recognition|journal = Electronics Letters|pages = 1370–1372|volume = 51|issue = 17|doi = 10.1049/el.2015.1428|first1 = A.K.|last1 = Maan|first2 = A.P.|last2 = James|first3 = S.|last3 = Dimitrijev|year = 2015|bibcode = 2015ElL....51.1370M|hdl = 10072/140989|s2cid = 61454815|url = https://semanticscholar.org/paper/48d3ab11ec6e213b62f11eedcfb7b7febb058674|hdl-access = free}}</ref> [[face recognition]]<ref>{{Cite journal|title = Memristive Threshold Logic Face Recognition|journal = Procedia Computer Science|date = 2014-01-01|pages = 98–103|volume = 41|series = 5th Annual International Conference on Biologically Inspired Cognitive Architectures, 2014 BICA|doi = 10.1016/j.procs.2014.11.090|first1 = Akshay Kumar|last1 = Maan|first2 = Dinesh S.|last2 = Kumar|first3 = Alex Pappachen|last3 = James|doi-access = free}}</ref> and [[object recognition]].<ref>{{Cite journal|title = Memristive Threshold Logic Circuit Design of Fast Moving Object Detection|journal = IEEE Transactions on Very Large Scale Integration (VLSI) Systems|date = 2015-10-01|issn = 1063-8210|pages = 2337–2341|volume = 23|issue = 10|doi = 10.1109/TVLSI.2014.2359801|first1 = A.K.|last1 = Maan|first2 = D.S.|last2 = Kumar|first3 = S.|last3 = Sugathan|first4 = A.P.|last4 = James|arxiv = 1410.1267|s2cid = 9647290}}</ref> They also find applications in replacing conventional digital logic gates.<ref>{{Cite journal|title = Resistive Threshold Logic|journal = IEEE Transactions on Very Large Scale Integration (VLSI) Systems|date = 2014-01-01|issn = 1063-8210|pages = 190–195|volume = 22|issue = 1|doi = 10.1109/TVLSI.2012.2232946|first1 = A.P.|last1 = James|first2 = L.R.V.J.|last2 = Francis|first3 = D.S.|last3 = Kumar|arxiv = 1308.0090|s2cid = 7357110}}</ref><ref>{{Cite journal|title = Threshold Logic Computing: Memristive-CMOS Circuits for Fast Fourier Transform and Vedic Multiplication|journal = IEEE Transactions on Very Large Scale Integration (VLSI) Systems|date = 2015-11-01|issn = 1063-8210|pages = 2690–2694|volume = 23|issue = 11|doi = 10.1109/TVLSI.2014.2371857|first1 = A.P.|last1 = James|first2 = D.S.|last2 = Kumar|first3 = A.|last3 = Ajayan|arxiv = 1411.5255|s2cid = 6076956}}</ref><br />
<br />
There exist several neuron inspired threshold logic functions implemented with memristors that have applications in high level pattern recognition applications. Some of the applications reported recently include speech recognition, face recognition and object recognition. They also find applications in replacing conventional digital logic gates.<br />
<br />
利用记忆电阻器实现的神经元启发阈值逻辑函数在高级模式识别中有着广泛的应用。最近报道的一些应用包括语音识别、人脸识别和物体识别。它们还可以用来取代传统的数字逻辑门。<br />
<br />
For ideal passive memristive circuits there is an exact equation (Caravelli-Traversa-[[Di Ventra]] equation) for the internal memory of the circuit:<ref>{{cite journal |last=Caravelli |display-authors=etal|arxiv=1608.08651 |title=The complex dynamics of memristive circuits: analytical results and universal slow relaxation |year=2017 |doi=10.1103/PhysRevE.95.022140 |pmid= 28297937 |volume=95 |issue= 2 |pages= 022140 |journal=Physical Review E|bibcode=2017PhRvE..95b2140C |s2cid=6758362}}</ref><br />
<br />
For ideal passive memristive circuits there is an exact equation (Caravelli-Traversa-Di Ventra equation) for the internal memory of the circuit:<br />
<br />
对于理想的无源记忆电路，电路的内部记忆有一个精确的方程(Caravelli-Traversa-Di Ventra 方程) :<br />
<br />
:<math> \frac{d}{dt} \vec{W} = \alpha \vec{W}-\frac{1}{\beta} (I+\xi \Omega W)^{-1} \Omega \vec S </math><br />
<br />
:\frac{d}{dt} \vec{W} = \alpha \vec{W}-\frac{1}{\beta} (I+\xi \Omega W)^{-1} \Omega \vec S<br />
<br />
:\frac{d}{dt} \vec{W} = \alpha \vec{W}-\frac{1}{\beta} (I+\xi \Omega W)^{-1} \Omega \vec S<br />
<br />
as a function of the properties of the physical memristive network and the external sources. In the equation above, <math>\alpha</math> is the "forgetting" time scale constant, <math> \xi=r-1</math> and <math>r=\frac{R_\text{off}}{R_\text{on}}</math> is the ratio of ''off'' and ''on'' values of the limit resistances of the memristors, <math> \vec S </math> is the vector of the sources of the circuit and <math>\Omega</math> is a projector on the fundamental loops of the circuit. The constant <math>\beta</math> has the dimension of a voltage and is associated to the properties of the [[memristor]]; its physical origin is the charge mobility in the conductor. The diagonal matrix and vector <math>W=\operatorname{diag}(\vec W)</math> and <math>\vec W</math> respectively, are instead the internal value of the memristors, with values between 0 and 1. This equation thus requires adding extra constraints on the memory values in order to be reliable.<br />
<br />
<nowiki>as a function of the properties of the physical memristive network and the external sources. In the equation above, \alpha is the "forgetting" time scale constant, \xi=r-1 and r=\frac{R_\text{off}}{R_\text{on}} is the ratio of off and on values of the limit resistances of the memristors, \vec S is the vector of the sources of the circuit and \Omega is a projector on the fundamental loops of the circuit. The constant \beta has the dimension of a voltage and is associated to the properties of the memristor; its physical origin is the charge mobility in the conductor. The diagonal matrix and vector W=\operatorname{diag}(\vec W) and \vec W respectively, are instead the internal value of the memristors, with values between 0 and 1. This equation thus requires adding extra constraints on the memory values in order to be reliable.</nowiki><br />
<br />
<nowiki>作为物理记忆网络和外部源的性质的函数。在上述方程中，α 是“遗忘”时间尺度常数，xi = r-1，r = frac { r _ text { off }{ on }{ r _ text { on }}是记忆电阻器极限电阻的开关和开关值之比，vec s 是电路源的矢量，Omega 是电路基本环路的投影仪。常数 β 具有电压的尺寸，与记忆电阻器的特性有关; 它的物理起源是导体中的电荷迁移率。对角矩阵和向量 w = 操作者名{ diag }(vec w)和 vec w 分别是记忆电阻器的内值，值在0到1之间。因此，这个等式需要在内存值上添加额外的约束，以保证可靠性。</nowiki><br />
<br />
==See also==<br />
== 相关词条 ==<br />
{{Columns-list|colwidth=18em|<br />
* [[AI accelerator (computer hardware)|AI accelerator]]<br />
* [[Artificial brain]]<br />
* [[Biomorphic]]<br />
* [[Cognitive computer]]<br />
* [[Computation and Neural Systems]]<br />
* [[Differentiable programming]]<br />
* [[Event camera]]<br />
* [[Neurorobotics]]<br />
* [[Optical flow sensor]]<br />
* [[Physical neural network]]<br />
* [[SpiNNaker]]<br />
* [[SyNAPSE]]<br />
* [[Retinomorphic sensor]]<br />
* [[Vision chip]]<br />
* [[Vision processing unit]]<br />
* [[Zeroth (software)]]<br />
* [[Hardware for artificial intelligence]]<br />
}}<br />
<br />
{{Portal bar|Electronics}}<br />
<br />
==References==<br />
{{Reflist|40em}}<br />
<br />
==External links==<br />
<!--======================== {{No more links}} ============================<br />
| PLEASE BE CAUTIOUS IN ADDING MORE LINKS TO THIS ARTICLE. Wikipedia |<br />
| is not a collection of links nor should it be used for advertising. |<br />
| |<br />
| Excessive or inappropriate links WILL BE DELETED. |<br />
| See [[Wikipedia:External links]] & [[Wikipedia:Spam]] for details. |<br />
| |<br />
| If there are already plentiful links, please propose additions or |<br />
| replacements on this article's discussion page, or submit your link |<br />
| to the relevant category at the Open Directory Project (dmoz.org) |<br />
| and link back to that category using the {{dmoz}} template. |<br />
======================={{No more links}}=============================--><br />
{{Commons category|Neuromorphic Engineering}}<br />
*[https://web.archive.org/web/20150727034331/http://ine-web.org/workshops/workshops-overview Telluride Neuromorphic Engineering Workshop]<br />
*[https://archive.today/20130115190057/http://capocaccia.ethz.ch/ CapoCaccia Cognitive Neuromorphic Engineering Workshop]<br />
*[https://web.archive.org/web/20190716132350/http://www.ine-web.org/ Institute of Neuromorphic Engineering]<br />
*[http://www.ine-news.org/ INE news site].<br />
*[http://www.frontiersin.org/neuromorphic_engineering Frontiers in Neuromorphic Engineering Journal]<br />
*[http://www.cns.caltech.edu/ Computation and Neural Systems] department at the [[California Institute of Technology]].<br />
*[http://www.humanbrainproject.eu/ Human Brain Project official site]<br />
*[https://www.the-scientist.com/features/building-a-silicon-brain-65738 Building a Silicon Brain:] Computer chips based on biological neurons may help simulate larger and more-complex brain models. May 1, 2019. SANDEEP RAVINDRAN<br />
<br />
<br />
*Telluride Neuromorphic Engineering Workshop<br />
*CapoCaccia Cognitive Neuromorphic Engineering Workshop<br />
* Institute of Neuromorphic Engineering<br />
*INE news site.<br />
*Frontiers in Neuromorphic Engineering Journal<br />
*Computation and Neural Systems department at the California Institute of Technology.<br />
*Human Brain Project official site<br />
*Building a Silicon Brain: Computer chips based on biological neurons may help simulate larger and more-complex brain models. May 1, 2019. SANDEEP RAVINDRAN<br />
<br />
=<nowiki>外部链接</nowiki>= <br />
*碲化物神经形态工程工作室<br />
*CapoCaccia 认知神经形态工程工作室<br />
*神经形态工程研究所<br />
*INE 新闻站点。<br />
*《神经形态工程学前沿》<br />
*加州理工学院计算与神经系统系。<br />
*人脑项目官方网站<br />
*打造硅脑: 基于生物神经元的计算机芯片可能有助于模拟更大、更复杂的大脑模型。2019年5月1日。SANDEEP RAVINDRAN<br />
<br />
{{Differentiable computing}}<br />
{{Authority control}}<br />
<br />
[[Category:Electrical engineering]]<br />
[[Category:Neuroscience]]<br />
[[分类:AI accelerators|*]]<br />
[[Category:Artificial intelligence]]<br />
[[Category:Robotics]]<br />
<br />
Category:Electrical engineering<br />
Category:Neuroscience<br />
*<br />
Category:Artificial intelligence<br />
Category:Robotics<br />
<br />
类别: 电气工程类别: 神经科学 * 类别: 人工智能类别: 机器人<br />
<br />
<noinclude><br />
<br />
<small>This page was moved from [[wikipedia:en:Neuromorphic engineering]]. Its edit history can be viewed at [[神经形态计算/edithistory]]</small></noinclude><br />
<br />
[[Category:待整理页面]]<br />
<br />
<references /></div>Shenky20https://wiki.swarma.org/index.php?title=%E7%A5%9E%E7%BB%8F%E5%BD%A2%E6%80%81%E8%AE%A1%E7%AE%97&diff=30183神经形态计算2022-04-14T06:52:30Z<p>Shenky20：/* 案例 */</p>
<hr />
<div><br />
此词条由神经动力学模型读书会词条梳理志愿者Shenky20翻译审校，未经专家审校，带来阅读不便，请见谅。<br />
<br />
{{Use American English|date = January 2019}}<br />
{{Use mdy dates|date = January 2019}}<br />
<br />
'''Neuromorphic engineering''', also known as '''neuromorphic computing''',<ref name=":3">{{Cite journal | doi = 10.1145/2601069| title = Neuromorphic computing gets ready for the (really) big time| journal = [[Communications of the ACM]]| volume = 57| issue = 6| year = 2014| last1 = Monroe | first1 = D. | pages = 13–15| s2cid = 20051102}}</ref><ref name=":4">{{Cite journal | doi = 10.1088/0957-4484/21/17/175202| pmid = 20368686| title = Nanotube devices based crossbar architecture: Toward neuromorphic computing| journal = Nanotechnology| volume = 21| issue = 17| pages = 175202| year = 2010| last1 = Zhao | first1 = W. S. | last2 = Agnus | first2 = G. | last3 = Derycke | first3 = V. | last4 = Filoramo | first4 = A. | last5 = Bourgoin | first5 = J. -P. | last6 = Gamrat | first6 = C. | bibcode = 2010Nanot..21q5202Z| url = https://zenodo.org/record/3428659}}</ref><ref name="humanbrainproject">{{YouTube|id=6RoiZ90mGfw|title=The Human Brain Project SP 9: Neuromorphic Computing Platform}}</ref> is the use of [[very-large-scale integration]] (VLSI) systems containing electronic [[analog circuit]]s to mimic neuro-biological architectures present in the nervous system. A neuromorphic computer/chip is any device that uses physical artificial neurons (made from silicon) to do computations.<ref name=":5">{{cite journal|last1=Mead|first1=Carver|title=Neuromorphic electronic systems|journal=Proceedings of the IEEE|date=1990|volume=78|issue=10|pages=1629–1636|doi=10.1109/5.58356|url=https://authors.library.caltech.edu/53090/1/00058356.pdf}}</ref><ref name=":2" /> In recent times, the term ''neuromorphic'' has been used to describe analog, digital, mixed-mode analog/digital VLSI, and software systems that implement models of [[neural system]]s (for [[perception]], [[motor control]], or [[multisensory integration]]). The implementation of neuromorphic computing on the hardware level can be realized by oxide-based [[memristor]]s,<ref name="Maan 1–13">{{Cite journal|last1=Maan|first1=A. K.|last2=Jayadevi|first2=D. A.|last3=James|first3=A. P.|date=2016-01-01|title=A Survey of Memristive Threshold Logic Circuits|journal=IEEE Transactions on Neural Networks and Learning Systems|volume=PP|issue=99|pages=1734–1746|doi=10.1109/TNNLS.2016.2547842|pmid=27164608|issn=2162-237X|arxiv=1604.07121|bibcode=2016arXiv160407121M|s2cid=1798273}}</ref> spintronic memories, threshold switches, and [[transistor]]s.<ref name=":6">{{Cite journal|title = Mott Memory and Neuromorphic Devices|journal = Proceedings of the IEEE|date = 2015-08-01|issn = 0018-9219|pages = 1289–1310|volume = 103|issue = 8|doi = 10.1109/JPROC.2015.2431914|first1 = You|last1 = Zhou|first2 = S.|last2 = Ramanathan|s2cid = 11347598|url = https://zenodo.org/record/895565}}</ref><ref name=":2">{{Cite document|title=Neuromorphic Circuits With Neural Modulation Enhancing the Information Content of Neural Signaling {{!}} International Conference on Neuromorphic Systems 2020|language=EN|doi=10.1145/3407197.3407204|s2cid=220794387}}</ref> Training software-based neuromorphic systems of spiking neural networks can be achieved using error backpropagation, e.g., using [[Python (programming language)|Python]] based frameworks such as snnTorch,<ref name=":7">{{cite journal |last1=Eshraghian|first1=Jason K.|last2=Ward|first2=Max|last3=Neftci |first3=Emre|last4=Wang|first4=Xinxin|last5=Lenz|first5=Gregor|last6=Dwivedi|first6=Girish|last7=Bennamoun|first7=Mohammed|last8=Jeong|first8=Doo Seok|last9=Lu|first9=Wei D.|title=Training Spiking Neural Networks Using Lessons from Deep Learning |date=1 October 2021 |arxiv=2109.12894 }}</ref> or using canonical learning rules from the biological learning literature, e.g., using BindsNet.<ref name=":8">{{Cite web | url=https://github.com/Hananel-Hazan/bindsnet | title=Hananel-Hazan/bindsnet: Simulation of spiking neural networks (SNNs) using PyTorch.| date=31 March 2020}}</ref><br />
<br />
<br />
'''<font color="#ff8000">神经形态工程Neuromorphic engineering</font>'''(也称为'''<font color="#ff8000">神经形态计算Neuromorphic computing</font>''')<ref name=":3" /><ref name=":4" /><ref name="humanbrainproject" />是指使用包含电子'''<font color="#ff8000">模拟电路Analog circuit</font>'''的'''<font color="#ff8000">超大规模集成电路Very-large-scale integration</font>'''系统来模拟神经系统中生理结构的研究方法。神经形态计算机或神经形态芯片包括任何使用由硅制成的人造神经元进行计算的设备。<ref name=":5" /><ref name=":2" />近年来，神经形态学（neuromorphic）这个术语被用来描述能够实现'''<font color="#ff8000">神经系统Neural system</font>'''模型功能(如'''<font color="#ff8000">感知Perception</font>'''、'''<font color="#ff8000">运动控制Motor control</font>'''，'''<font color="#ff8000">多感官整合Multisensory integration</font>'''等)的模拟、数字、模拟/数字混合模式超大规模集成电路和软件系统。神经形态计算的硬件实现可以通过基于氧化物的'''<font color="#ff8000">记忆电阻器Memristor</font>'''、<ref name="Maan 1–13" />自旋电子存储器、阈值开关和'''<font color="#ff8000">晶体管Transistor</font>'''来实现。<ref name=":6" /><ref name=":2" />对基于软件的脉冲神经网络系统的训练可以通过误差反向传播机制来实现，例如，使用snnTorch等基于Python的框架，<ref name=":7" />或使用BindsNet等典型的受生物启发的学习模式。<ref name=":8" /><br />
<br />
A key aspect of neuromorphic engineering is understanding how the morphology of individual neurons, circuits, applications, and overall architectures creates desirable computations, affects how information is represented, influences robustness to damage, incorporates learning and development, adapts to local change (plasticity), and facilitates evolutionary change.<br />
<br />
<br />
神经形态工程领域的一个关键问题，就是理解'''<font color="#32CD32">单个神经元形态、神经回路、应用和整体结构</font>'''如何产生理想的计算，如何影响信息的表示和对破坏的鲁棒性，如何整合学习和发展，如何适应局部变化(可塑性) 并促进逐渐发展的变化。<br />
<br />
Neuromorphic engineering is an interdisciplinary subject that takes inspiration from [[biology]], [[physics]], [[mathematics]], [[computer science]], and [[electronic engineering]]<ref name=":2" /> to design artificial neural systems, such as [[Machine vision|vision systems]], head-eye systems, auditory processors, and autonomous robots, whose physical architecture and design principles are based on those of biological nervous systems.<ref name=":9">{{Cite journal | doi = 10.1155/2012/705483| title = Qualitative Functional Decomposition Analysis of Evolved Neuromorphic Flight Controllers| journal = Applied Computational Intelligence and Soft Computing| volume = 2012| pages = 1–21| year = 2012| last1 = Boddhu | first1 = S. K. | last2 = Gallagher | first2 = J. C. | doi-access = free}}</ref> It was developed by [[Carver Mead]]<ref name=":10">{{cite web|title=carver mead website|last1=Mead|first1=Carver|url=http://www.carvermead.caltech.edu/index.html|website=carvermead}}</ref> in the late 1980s.<br />
<br />
<br />
神经形态工程是以'''<font color="#ff8000">生物学Biology</font>'''、'''<font color="#ff8000">物理学Physics</font>'''、'''<font color="#ff8000">数学Mathematics</font>'''、'''<font color="#ff8000">计算机科学Computer science</font>'''和'''<font color="#ff8000">电子工程Electronic engineering<ref name=":2" /></font>'''等学科为基础，设计人工神经系统（如'''<font color="#ff8000">视觉系统Vision systems</font>'''、头眼系统、听觉处理器以及物理结构和设计原则都受启发于生物神经系统的自主机器人）的一门交叉学科。<ref name=":9" />20世纪80年代后期，卡弗·米德极大地推动了神经形态工程领域的发展。<ref name=":10" /><br />
<br />
==Neurological inspiration==<br />
==神经生物学启发==<br />
Neuromorphic engineering is set apart by the inspiration it takes from what we know about the structure and operations of the [[brain]]. Neuromorphic engineering translates what we know about the brain's function into computer systems. Work has mostly focused on replicating the analog nature of [[biological computation]] and the role of [[neuron]]s in [[cognition]].<br />
<br />
<br />
神经形态工程形成的灵感来源于现已知的'''<font color="#ff8000">大脑Brain</font>'''结构及运作机制知识，它将我们对大脑功能的了解用于研究和优化计算机系统。该领域工作主要集中于对'''<font color="#ff8000">生物计算Biological computation</font>'''模拟特性和'''<font color="#ff8000">神经元Neuron</font>'''在'''<font color="#ff8000">认知Cognition</font>'''中所发挥作用的复刻。<br />
<br />
The biological processes of neurons and their [[synapse]]s are dauntingly complex, and thus very difficult to artificially simulate. A key feature of biological brains is that all of the processing in neurons use analog chemical signals. This makes it hard to replicate brains in computers because the current generation of computers is completely digital. However, the characteristics of these parts can be abstracted into mathematical functions that closely capture the essence of the neuron's operations.<br />
<br />
神经元及其'''<font color="#ff8000">突触Synapse</font>'''的生理学过程极其复杂，难以进行人工模拟。神经元在所有处理过程中都使用化学模拟信号，这是大脑的一个关键生理学特征。这个特征大大增加了在计算机中复制大脑的难度，因为目前的计算机是完全数字化的。然而，部分特征可以抽象为数学函数，这些函数能够紧密捕捉神经元运行的本质。<br />
<br />
The goal of neuromorphic computing is not to perfectly mimic the brain and all of its functions, but instead to extract what is known of its structure and operations to be used in a practical computing system. No neuromorphic system will claim nor attempt to reproduce every element of neurons and synapses, but all adhere to the idea that computation is highly [[distributed processing|distributed]] throughout a series of small computing elements analogous to a neuron. While this sentiment is standard, researchers chase this goal with different methods.<ref name=":11">{{Cite journal | doi = 10.1088/1741-2560/13/5/051001| title = Large-scale neuromorphic computing systems| journal = Journal of Neural Engineering| volume = 13| pages = 1–15| year = 2016| last1 = Furber | first1 = Steve| issue = 5| pmid = 27529195| bibcode = 2016JNEng..13e1001F| doi-access = free}}</ref><br />
<br />
神经形态计算的目标不是完美地模拟大脑及其所有功能，而是利用已知的大脑结构和运转机制来研发或优化实际的计算系统。任何神经形态学系统都不会声称或试图复制神经元和突触中的每一个元素，但所有人都一致认可将计算高度'''<font color="#ff8000">分散Distribute</font>'''于一系列类似于神经元的小型计算元素的理念。研究人员用不同的方法来追求这一普遍目标。<ref name=":11" /><br />
<br />
<br />
==Examples==<br />
==案例==<br />
As early as 2006, researchers at [[Georgia Tech]] published a field programmable neural array.<ref name=":12">{{Cite book|title = A field programmable neural array|last1 = Farquhar|first1 = Ethan|date = May 2006|journal = IEEE International Symposium on Circuits and Systems|pages = 4114–4117|last2 = Hasler|first2 = Paul.|doi = 10.1109/ISCAS.2006.1693534|isbn = 978-0-7803-9389-9|s2cid = 206966013}}</ref> This chip was the first in a line of increasingly complex arrays of floating gate transistors that allowed programmability of charge on the gates of [[MOSFET]]s to model the channel-ion characteristics of neurons in the brain and was one of the first cases of a silicon programmable array of neurons.<br />
<br />
早在2006年，佐治亚理工学院的研究人员就研发出了一种现场可编程神经阵列。<ref name=":12" />在此之后，一系列越来越复杂的浮栅晶体管阵列被成功研发出来，这些晶体管阵列可以通过在'''<font color="#ff8000">金属-氧化物半导体效应晶体管MOSFET</font>'''的栅极上编程来模拟大脑中神经元的离子通道特性，这也是以硅为主要材料的可编程神经元阵列的最早成功案例之一。<br />
<br />
In November 2011, a group of [[MIT]] researchers created a computer chip that mimics the analog, ion-based communication in a synapse between two neurons using 400 transistors and standard [[CMOS]] manufacturing techniques.<ref name=":13">{{cite web|title=MIT creates "brain chip"|url=http://www.extremetech.com/extreme/105067-mit-creates-brain-chip|access-date=4 December 2012}}</ref><ref name="Neuromorphic silicon paper">{{cite journal|title=Neuromorphic silicon neurons and large-scale neural networks: challenges and opportunities|doi=10.3389/fnins.2011.00108|pmid=21991244|pmc=3181466|volume=5|pages=108|journal=Frontiers in Neuroscience|year=2011|last1=Poon|first1=Chi-Sang|last2=Zhou|first2=Kuan|doi-access=free}}</ref><br />
<br />
2011年11月，麻省理工学院的一组研究人员研发出一种计算机芯片，该芯片上使用标准的'''<font color="#ff8000">互补金属氧化物半导体CMOS</font>'''制造技术集成了400个晶体管，能够模拟神经元间突触中基于离子的通讯。<ref name=":13" /><ref name="Neuromorphic silicon paper" /><br />
<br />
In June 2012, [[spintronic]] researchers at [[Purdue University]] presented a paper on the design of a neuromorphic chip using [[Spin valve|lateral spin valve]]s and [[memristor]]s. They argue that the architecture works similarly to neurons and can therefore be used to test methods of reproducing the brain's processing. In addition, these chips are significantly more energy-efficient than conventional ones.<ref name="Spin Devices Prop">{{Cite arXiv|title=Proposal For Neuromorphic Hardware Using Spin Devices|eprint=1206.3227|last1=Sharad|first1=Mrigank|last2=Augustine|first2=Charles|last3=Panagopoulos|first3=Georgios|last4=Roy|first4=Kaushik|class=cond-mat.dis-nn|year=2012}}</ref><br />
<br />
2012年6月，普渡大学的'''<font color="#ff8000">自旋电子学Spintronic</font>'''研究人员发表了一篇关于利用'''<font color="#ff8000">侧向自旋阀Lateral spin valves</font>'''和'''<font color="#ff8000">记忆电阻器Memristors</font>'''设计神经形态芯片的论文。他们认为，这种芯片结构的工作原理与神经元相似，因此可以用于大脑运行机制的复刻方法的测试。此外，这些芯片在能耗方面明显优于传统芯片。<ref name="Spin Devices Prop" /><br />
<br />
Research at [[HP Labs]] on Mott memristors has shown that while they can be non-[[Volatile memory|volatile]], the volatile behavior exhibited at temperatures significantly below the [[phase transition]] temperature can be exploited to fabricate a [[neuristor]],<ref name=":0" /> a biologically-inspired device that mimics behavior found in neurons.<ref name=":0">{{Cite journal | doi = 10.1038/nmat3510| pmid = 23241533| title = A scalable neuristor built with Mott memristors| journal = Nature Materials| volume = 12| issue = 2| pages = 114–7| year = 2012| last1 = Pickett | first1 = M. D. | last2 = Medeiros-Ribeiro | first2 = G. | last3 = Williams | first3 = R. S. | bibcode = 2013NatMa..12..114P| s2cid = 16271627| url = https://semanticscholar.org/paper/b6ba6f496ace2b947f111059663e76bb60e9efeb}}</ref> In September 2013, they presented models and simulations that show how the spiking behavior of these neuristors can be used to form the components required for a [[Turing machine]].<ref name=":14">{{cite journal|doi=10.1088/0957-4484/24/38/384002|title=Phase transitions enable computational universality in neuristor-based cellular automata|author1=Matthew D Pickett|author2=R Stanley Williams|name-list-style=amp|date=September 2013|publisher=IOP Publishing Ltd|journal=Nanotechnology|volume=24|issue=38|pmid=23999059|bibcode=2013Nanot..24L4002P|at=384002}}</ref><br />
<br />
'''<font color="#ff8000">惠普实验室HP labs</font>'''在莫特记忆电阻器上的研究表明，尽管它们可以是非'''<font color="#ff8000">易失性Volatile</font>'''的，但是在'''<font color="#ff8000">相变Phase transition</font>'''温度以下时表现出的易失性行为可以被用来制造'''<font color="#ff8000">类神经元电阻器Neuristor</font>'''(一种生物学启发的模仿神经元行为的硬件)<ref name=":0" />。2013年9月，他们通过模型和仿真展示了这些类神经元电阻器的脉冲行为如何产生'''<font color="#ff8000">图灵机Turing machine</font>'''的所需元素。<ref name=":14" /><br />
<br />
[[Neurogrid]], built by ''Brains in Silicon'' at [[Stanford University]],<ref name=":15">{{cite journal|last1=Boahen|first1=Kwabena|title=Neurogrid: A Mixed-Analog-Digital Multichip System for Large-Scale Neural Simulations|journal=Proceedings of the IEEE|date=24 April 2014|volume=102|issue=5|pages=699–716|doi=10.1109/JPROC.2014.2313565|s2cid=17176371}}</ref> is an example of hardware designed using neuromorphic engineering principles. The circuit board is composed of 16 custom-designed chips, referred to as NeuroCores. Each NeuroCore's analog circuitry is designed to emulate neural elements for 65536 neurons, maximizing energy efficiency. The emulated neurons are connected using digital circuitry designed to maximize spiking throughput.<ref name=":16">{{cite journal|doi=10.1038/503022a|pmid = 24201264|title = Neuroelectronics: Smart connections|journal = Nature|volume = 503|issue = 7474|pages = 22–4|year = 2013|last1 = Waldrop|first1 = M. Mitchell|bibcode = 2013Natur.503...22W|doi-access = free}}</ref><ref name=":17">{{cite journal|doi=10.1109/JPROC.2014.2313565|title = Neurogrid: A Mixed-Analog-Digital Multichip System for Large-Scale Neural Simulations|journal = Proceedings of the IEEE|volume = 102|issue = 5|pages = 699–716|year = 2014|last1 = Benjamin|first1 = Ben Varkey|last2 = Peiran Gao|last3 = McQuinn|first3 = Emmett|last4 = Choudhary|first4 = Swadesh|last5 = Chandrasekaran|first5 = Anand R.|last6 = Bussat|first6 = Jean-Marie|last7 = Alvarez-Icaza|first7 = Rodrigo|last8 = Arthur|first8 = John V.|last9 = Merolla|first9 = Paul A.|last10 = Boahen|first10 = Kwabena|s2cid = 17176371}}</ref><br />
<br />
'''<font color="#ff8000">神经栅格Neurogrid<ref name=":15" /></font>'''是由斯坦福大学Brains in Silicon公司研发的、使用神经形态工程原理设计的硬件。该电路板由16个定制设计的芯片组成(NeuroCores)。在设计中，每个NeuroCore芯片的模拟电路对65536个神经元的神经元素进行模拟，以最大限度地提高能量效率。模拟出的神经元通过设计的数字电路连接，以最大化脉冲吞吐量。<ref name=":16" /><ref name=":17" /><br />
<br />
A research project with implications for neuromorphic engineering is the [[Human Brain Project]] that is attempting to simulate a complete human brain in a supercomputer using biological data. It is made up of a group of researchers in neuroscience, medicine, and computing.<ref>{{cite web|title=Involved Organizations|url=http://www.humanbrainproject.eu/partners.html|access-date=22 February 2013|url-status=dead|archive-url=https://web.archive.org/web/20130302142627/http://www.humanbrainproject.eu/partners.html|archive-date=2 March 2013}}</ref> [[Henry Markram]], the project's co-director, has stated that the project proposes to establish a foundation to explore and understand the brain and its diseases, and to use that knowledge to build new computing technologies. The three primary goals of the project are to better understand how the pieces of the brain fit and work together, to understand how to objectively diagnose and treat brain diseases, and to use the understanding of the human brain to develop neuromorphic computers. That the simulation of a complete human brain will require a supercomputer a thousand times more powerful than today's encourages the current focus on neuromorphic computers.<ref>{{cite web|title=Human Brain Project|url=http://www.humanbrainproject.eu|access-date=22 February 2013}}</ref> $1.3 billion has been allocated to the project by The [[European Commission]].<ref>{{cite web|title=The Human Brain Project and Recruiting More Cyberwarriors|url=http://www.marketplace.org/topics/tech/human-brain-project-and-recruiting-more-cyberwarriors|access-date=22 February 2013|date=January 29, 2013}}</ref><br />
<br />
A research project with implications for neuromorphic engineering is the Human Brain Project that is attempting to simulate a complete human brain in a supercomputer using biological data. It is made up of a group of researchers in neuroscience, medicine, and computing. Henry Markram, the project's co-director, has stated that the project proposes to establish a foundation to explore and understand the brain and its diseases, and to use that knowledge to build new computing technologies. The three primary goals of the project are to better understand how the pieces of the brain fit and work together, to understand how to objectively diagnose and treat brain diseases, and to use the understanding of the human brain to develop neuromorphic computers. That the simulation of a complete human brain will require a supercomputer a thousand times more powerful than today's encourages the current focus on neuromorphic computers. $1.3 billion has been allocated to the project by The European Commission.<br />
<br />
一个对神经形态工程有影响的研究项目是人脑项目，它试图用生物数据在超级计算机中模拟完整的人脑。它由一群神经科学、医学和计算机科学的研究人员组成。该项目的联合主管亨利•马克拉姆(Henry Markram)表示，该项目提议建立一个基础，以探索和了解大脑及其疾病，并利用这些知识构建新的计算机技术。这个项目的三个主要目标是: 更好地理解大脑的各个部分是如何相互匹配和协同工作的; 理解如何客观地诊断和治疗脑部疾病; 以及利用对人类大脑的理解来开发神经形态计算机。模拟一个完整的人类大脑需要一台比现在强大一千倍的超级计算机，这鼓励了当前对神经形态计算机的关注。欧洲委员会已经拨款13亿美元用于这个项目。<br />
<br />
Other research with implications for neuromorphic engineering involves the [[BRAIN Initiative]]<ref name="economist">[https://www.economist.com/news/science-and-technology/21582495-computers-will-help-people-understand-brains-better-and-understanding-brains Neuromorphic computing: The machine of a new soul], The Economist, 2013-08-03</ref> and the [[TrueNorth]] chip from [[IBM]].<ref>{{cite journal|last1=Modha|first1=Dharmendra|title=A million spiking-neuron integrated circuit with a scalable communication network and interface|journal=Science|date=Aug 2014|volume=345|issue=6197|pages=668–673|doi=10.1126/science.1254642|pmid=25104385|bibcode=2014Sci...345..668M|s2cid=12706847}}</ref> Neuromorphic devices have also been demonstrated using nanocrystals, nanowires, and conducting polymers.<ref>{{Cite web|url=http://jessamynfairfield.com/wp-content/uploads/2017/03/PWMar17Fairfield.pdf|title=Smarter Machines|last=Fairfield|first=Jessamyn|date=March 1, 2017}}</ref><br />
<br />
Other research with implications for neuromorphic engineering involves the BRAIN InitiativeNeuromorphic computing: The machine of a new soul, The Economist, 2013-08-03 and the TrueNorth chip from IBM. Neuromorphic devices have also been demonstrated using nanocrystals, nanowires, and conducting polymers.<br />
<br />
其他与神经形态工程有关的研究还包括 BRAIN initiativity/euromorphic computing: The machine of a new soul，The Economist，2013-08-03 and The TrueNorth chip from IBM。神经形态学设备也已经被证明使用纳米晶体、纳米线和导电聚合物。<br />
<br />
[[Intel]] unveiled its neuromorphic research chip, called “[[Intel Loihi|Loihi]]”, in October 2017. The chip uses an asynchronous [[spiking neural network]] (SNN) to implement adaptive self-modifying event-driven fine-grained parallel computations used to implement learning and inference with high efficiency.<ref>{{cite journal |last1=Davies |first1=Mike |title=Loihi: A Neuromorphic Manycore Processor with On-Chip Learning |journal=IEEE Micro |date=January 16, 2018 |volume=38 |issue=1 |pages=82–99 |display-authors=etal|doi=10.1109/MM.2018.112130359 |s2cid=3608458 }}</ref><ref name="Morris2017">{{cite web |last1=Morris |first1=John |title=Why Intel built a neuromorphic chip |url=https://www.zdnet.com/article/why-intel-built-a-neuromorphic-chip/ |website=ZDNet |access-date=17 August 2018 |language=en}}</ref><br />
<br />
Intel unveiled its neuromorphic research chip, called “Loihi”, in October 2017. The chip uses an asynchronous spiking neural network (SNN) to implement adaptive self-modifying event-driven fine-grained parallel computations used to implement learning and inference with high efficiency.<br />
<br />
2017年10月，英特尔公布了其神经形态研究芯片“ Loihi”。该芯片采用异步脉冲神经网络(SNN)实现自适应自修改事件驱动的细粒度并行计算，实现了高效的学习和推理。<br />
<br />
[[IMEC]], a Belgium-based nanoelectronics research center, demonstrated the world's first self-learning neuromorphic chip. The brain-inspired chip, based on OxRAM technology, has the capability of self-learning and has been demonstrated to have the ability to compose music.<ref>{{cite web |title=Imec demonstrates self-learning neuromorphic chip that composes music |url=https://www.imec-int.com/en/articles/imec-demonstrates-self-learning-neuromorphic-chip-that-composes-music |website=IMEC International |access-date=1 October 2019}}</ref> IMEC released the 30-second tune composed by the prototype. The chip was sequentially loaded with songs in the same time signature and style. The songs were old Belgian and French flute minuets, from which the chip learned the rules at play and then applied them.<ref>{{cite web|last1=Bourzac|first1=Katherine|title=A Neuromorphic Chip That Makes Music|url=https://spectrum.ieee.org/a-neuromorphic-chip-that-makes-music|url-status=live|access-date=1 October 2019|website=IEEE Spectrum|date=May 23, 2017}}</ref><br />
<br />
IMEC, a Belgium-based nanoelectronics research center, demonstrated the world's first self-learning neuromorphic chip. The brain-inspired chip, based on OxRAM technology, has the capability of self-learning and has been demonstrated to have the ability to compose music. IMEC released the 30-second tune composed by the prototype. The chip was sequentially loaded with songs in the same time signature and style. The songs were old Belgian and French flute minuets, from which the chip learned the rules at play and then applied them.<br />
<br />
比利时的一个纳米电子学研究中心 IMEC 展示了世界上第一个自学习神经形态芯片。这种基于 OxRAM 技术的大脑启发芯片具有自学习能力，并已被证明具有创作音乐的能力。IMEC 发布了由原型机谱写的30秒曲调。芯片按顺序加载同时签名和风格的歌曲。歌曲是古老的比利时和法国长笛小步舞曲，筹码从中学习游戏规则，然后应用它们。<br />
<br />
[[Blue Brain Project|The Blue Brain Project]], led by Henry Markram, aims to build biologically detailed digital reconstructions and simulations of the mouse brain. The Blue Brain Project has created in silico models of rodent brains, while attempting to replicate as many details about its biology as possible. The supercomputer-based simulations offer new perspectives on understanding the structure and functions of the brain.<br />
<br />
The Blue Brain Project, led by Henry Markram, aims to build biologically detailed digital reconstructions and simulations of the mouse brain. The Blue Brain Project has created in silico models of rodent brains, while attempting to replicate as many details about its biology as possible. The supercomputer-based simulations offer new perspectives on understanding the structure and functions of the brain.<br />
<br />
由亨利 · 马克拉姆领导的蓝脑计划旨在建立小鼠大脑的生物细节数字重建和模拟。蓝色大脑项目已经在啮齿动物大脑的电子模型中创建，同时试图尽可能多地复制其生物学细节。基于超级计算机的模拟为理解大脑的结构和功能提供了新的视角。<br />
<br />
The European Union funded a series of projects at the University of Heidelberg, which led to the development of <br />
[[BrainScaleS]] (brain-inspired multiscale computation in neuromorphic hybrid systems), a hybrid analog [[neuromorphic]] supercomputer located at Heidelberg University, Germany. It was developed as part of the [[Human Brain Project]] neuromorphic computing platform and is the complement to the [[SpiNNaker]] supercomputer (which is based on digital technology). The architecture used in BrainScaleS mimics biological neurons and their connections on a physical level; additionally, since the components are made of silicon, these model neurons operate on average 864 times (24 hours of real time is 100 seconds in the machine simulation) that of their biological counterparts.<ref>{{Cite web|date=2016-03-21|title=Beyond von Neumann, Neuromorphic Computing Steadily Advances|url=https://www.hpcwire.com/2016/03/21/lacking-breakthrough-neuromorphic-computing-steadily-advance/|access-date=2021-10-08|website=HPCwire|language=en-US}}</ref><br />
<br />
The European Union funded a series of projects at the University of Heidelberg, which led to the development of <br />
BrainScaleS (brain-inspired multiscale computation in neuromorphic hybrid systems), a hybrid analog neuromorphic supercomputer located at Heidelberg University, Germany. It was developed as part of the Human Brain Project neuromorphic computing platform and is the complement to the SpiNNaker supercomputer (which is based on digital technology). The architecture used in BrainScaleS mimics biological neurons and their connections on a physical level; additionally, since the components are made of silicon, these model neurons operate on average 864 times (24 hours of real time is 100 seconds in the machine simulation) that of their biological counterparts.<br />
<br />
欧盟资助了海德堡大学的一系列项目，这些项目导致了 BrainScaleS (神经形态混合系统中受大脑启发的多尺度计算)的发展，这是一台位于德国海德堡大学的混合模拟神经形态超级计算机。它是作为人脑计划神经形态计算平台的一部分而开发的，是 SpiNNaker 超级计算机(基于数字技术)的补充。大脑尺度中使用的体系结构模拟了生物神经元及其在物理层面上的连接; 此外，由于这些组件是由硅制成的，这些模型神经元平均运行864倍(在机器模拟中，24小时的实时时间是100秒)。<br />
<br />
===Neuromorphic sensors ===<br />
===神经形态传感器===<br />
The concept of neuromorphic systems can be extended to sensors (not just to computation). An example of this applied to detecting [[light]] is the [[retinomorphic sensor]] or, when employed in an array, the [[event camera]].<br />
<br />
The concept of neuromorphic systems can be extended to sensors (not just to computation). An example of this applied to detecting light is the retinomorphic sensor or, when employed in an array, the event camera.<br />
<br />
神经形态系统的概念可以扩展到传感器(而不仅仅是计算)。用于检测光线的一个例子是视网膜变形传感器，或者在阵列中使用的事件摄像机。<br />
<br />
<br />
==Ethical considerations==<br />
==伦理问题==<br />
While the interdisciplinary concept of neuromorphic engineering is relatively new, many of the same ethical considerations apply to neuromorphic systems as apply to [[human-like machines]] and [[artificial intelligence]] in general. However, the fact that neuromorphic systems are designed to mimic a [[human brain]] gives rise to unique ethical questions surrounding their usage.<br />
<br />
While the interdisciplinary concept of neuromorphic engineering is relatively new, many of the same ethical considerations apply to neuromorphic systems as apply to human-like machines and artificial intelligence in general. However, the fact that neuromorphic systems are designed to mimic a human brain gives rise to unique ethical questions surrounding their usage.<br />
<br />
虽然神经形态工程的跨学科概念相对较新，但许多同样的伦理考虑适用于神经形态系统，就像适用于类人机器和一般人工智能一样。然而，神经形态系统是为了模仿人类大脑而设计的这一事实引起了围绕其使用的独特的伦理问题。<br />
<br />
However, the practical debate is that neuromorphic hardware as well as artificial "neural networks" are immensely simplified models of how the brain operates or processes information at a much lower [[complex system|complexity in terms of size and functional technology]] and a much more regular structure in terms of [[brain connectivity|connectivity]]. Comparing [[neuromorphic chip]]s to the brain is a very crude comparison similar to comparing a plane to a bird just because they both have wings and a tail. The fact is that neural cognitive systems are many orders of magnitude more [[Energy efficiency (physics)|energy-]] and compute-efficient than current state-of-the-art AI and neuromorphic engineering is an attempt to narrow this gap by inspiring from the brain's mechanism just like many engineering designs have [[bioengineering|bio-inspired features]].<br />
<br />
However, the practical debate is that neuromorphic hardware as well as artificial "neural networks" are immensely simplified models of how the brain operates or processes information at a much lower complexity in terms of size and functional technology and a much more regular structure in terms of connectivity. Comparing neuromorphic chips to the brain is a very crude comparison similar to comparing a plane to a bird just because they both have wings and a tail. The fact is that neural cognitive systems are many orders of magnitude more energy- and compute-efficient than current state-of-the-art AI and neuromorphic engineering is an attempt to narrow this gap by inspiring from the brain's mechanism just like many engineering designs have bio-inspired features.<br />
<br />
然而，实际上的争论是，神经形态硬件和人工“神经网络”是大脑如何运作或处理信息的极其简化的模型，在大小和功能技术方面的复杂性要低得多，在连接方面的结构则更加规则。将神经形态芯片与大脑进行比较是一种非常粗糙的比较，类似于仅仅因为一架飞机有翅膀和一条尾巴就将它与一只鸟进行比较。事实上，神经认知系统比当前最先进的人工智能具有更多的能量和计算效率，而神经形态工程是一种通过从大脑机制中激发灵感来缩小这种差距的尝试，就像许多工程设计具有生物启发的特征一样。数量级。<br />
<br />
===Democratic concerns ===<br />
===公众担忧===<br />
Significant ethical limitations may be placed on neuromorphic engineering due to public perception.<ref>{{Cite report|url=https://ai100.stanford.edu/sites/g/files/sbiybj9861/f/ai_100_report_0831fnl.pdf|title=Artificial Intelligence and Life in 2030|author=2015 Study Panel|date=September 2016|work=One Hundred Year Study on Artificial Intelligence (AI100)|publisher=Stanford University}}</ref> Special [[Eurobarometer]] 382: Public Attitudes Towards Robots, a survey conducted by the European Commission, found that 60% of [[European Union]] citizens wanted a ban of robots in the care of children, the elderly, or the disabled. Furthermore, 34% were in favor of a ban on robots in education, 27% in healthcare, and 20% in leisure. The European Commission classifies these areas as notably “human.” The report cites increased public concern with robots that are able to mimic or replicate human functions. Neuromorphic engineering, by definition, is designed to replicate the function of the human brain.<ref name=":1">{{Cite web|url=http://ec.europa.eu/commfrontoffice/publicopinion/archives/ebs/ebs_382_en.pdf|title=Special Eurobarometer 382: Public Attitudes Towards Robots|last=European Commission|date=September 2012|website=European Commission}}</ref><br />
<br />
Significant ethical limitations may be placed on neuromorphic engineering due to public perception. Special Eurobarometer 382: Public Attitudes Towards Robots, a survey conducted by the European Commission, found that 60% of European Union citizens wanted a ban of robots in the care of children, the elderly, or the disabled. Furthermore, 34% were in favor of a ban on robots in education, 27% in healthcare, and 20% in leisure. The European Commission classifies these areas as notably “human.” The report cites increased public concern with robots that are able to mimic or replicate human functions. Neuromorphic engineering, by definition, is designed to replicate the function of the human brain.<br />
<br />
由于公众的认知，神经形态工程学可能会受到严重的伦理限制。欧盟委员会进行的一项调查发现，60% 的欧盟公民希望禁止照顾儿童、老人或残疾人的机器人。此外，34% 的人支持禁止机器人用于教育，27% 的人支持医疗保健，20% 的人支持休闲。欧盟委员会将这些地区明显归类为“人类”报告指出，公众越来越关注能够模仿或复制人类功能的机器人。神经形态工程，顾名思义，是为了复制人脑的功能而设计的。<br />
<br />
The democratic concerns surrounding neuromorphic engineering are likely to become even more profound in the future. The European Commission found that EU citizens between the ages of 15 and 24 are more likely to think of robots as human-like (as opposed to instrument-like) than EU citizens over the age of 55. When presented an image of a robot that had been defined as human-like, 75% of EU citizens aged 15–24 said it corresponded with the idea they had of robots while only 57% of EU citizens over the age of 55 responded the same way. The human-like nature of neuromorphic systems, therefore, could place them in the categories of robots many EU citizens would like to see banned in the future.<ref name=":1" /><br />
<br />
The democratic concerns surrounding neuromorphic engineering are likely to become even more profound in the future. The European Commission found that EU citizens between the ages of 15 and 24 are more likely to think of robots as human-like (as opposed to instrument-like) than EU citizens over the age of 55. When presented an image of a robot that had been defined as human-like, 75% of EU citizens aged 15–24 said it corresponded with the idea they had of robots while only 57% of EU citizens over the age of 55 responded the same way. The human-like nature of neuromorphic systems, therefore, could place them in the categories of robots many EU citizens would like to see banned in the future.<br />
<br />
围绕神经形态工程的民主关注可能在未来变得更加深刻。欧盟委员会(European Commission)发现，15至24岁的欧盟公民比55岁以上的欧盟公民更有可能认为机器人像人(而不是像仪器)。当看到一张被定义为“类人”的机器人图片时，年龄在15岁至24岁之间的欧盟公民中有75% 的人表示，这与他们对机器人的想法相符，而55岁以上的欧盟公民中只有57% 的人有同样的反应。因此，类似人类的神经形态系统，可以把它们归入许多欧盟公民希望在未来禁止使用的机器人类别。<br />
<br />
===Personhood===<br />
=== 人格问题===<br />
As neuromorphic systems have become increasingly advanced, some scholars{{who|date=August 2021}} have advocated for granting [[personhood]] rights to these systems. If the brain is what grants humans their personhood, to what extent does a neuromorphic system have to mimic the human brain to be granted personhood rights? Critics of technology development in the [[Human Brain Project]], which aims to advance brain-inspired computing, have argued that advancement in neuromorphic computing could lead to machine consciousness or personhood.<ref>{{Cite journal|last=Aicardi|first=Christine|date=September 2018|title=Accompanying technology development in the Human Brain Project: From foresight to ethics management|journal=Futures|volume=102|pages=114–124|doi=10.1016/j.futures.2018.01.005|doi-access=free}}</ref> If these systems are to be treated as people, critics argue, then many tasks humans perform using neuromorphic systems, including the act of termination of neuromorphic systems, may be morally impermissible as these acts would violate the autonomy of the neuromorphic systems.<ref>{{Cite journal|last=Lim|first=Daniel|date=2014-06-01|title=Brain simulation and personhood: a concern with the Human Brain Project|journal=Ethics and Information Technology|language=en|volume=16|issue=2|pages=77–89|doi=10.1007/s10676-013-9330-5|s2cid=17415814|issn=1572-8439}}</ref><br />
<br />
As neuromorphic systems have become increasingly advanced, some scholars have advocated for granting personhood rights to these systems. If the brain is what grants humans their personhood, to what extent does a neuromorphic system have to mimic the human brain to be granted personhood rights? Critics of technology development in the Human Brain Project, which aims to advance brain-inspired computing, have argued that advancement in neuromorphic computing could lead to machine consciousness or personhood. If these systems are to be treated as people, critics argue, then many tasks humans perform using neuromorphic systems, including the act of termination of neuromorphic systems, may be morally impermissible as these acts would violate the autonomy of the neuromorphic systems.<br />
<br />
随着神经形态系统的日益发展，一些学者主张赋予这些系统人格权。如果是大脑赋予了人类人格，那么神经形态系统在多大程度上必须模仿人类大脑才能被赋予人格权利？“人脑计划”旨在推进以大脑为灵感的计算机技术发展，该计划的批评者认为，神经形态计算机技术的进步可能导致机器意识或人格的形成。批评家认为，如果这些系统被当作人来对待，那么人类使用神经形态系统执行的许多任务，包括终止神经形态系统的行为，可能在道德上是不允许的，因为这些行为将违反神经形态系统的自主性。<br />
<br />
==Dual use (military applications)==<br />
==军民两用技术==<br />
The [[Joint Artificial Intelligence Center]], a branch of the U.S. military, is a center dedicated to the procurement and implementation of AI software and neuromorphic hardware for combat use. Specific applications include smart headsets/goggles and robots. JAIC intends to rely heavily on neuromorphic technology to connect "every fighter every shooter" within a network of neuromorphic-enabled units.<br />
<br />
The Joint Artificial Intelligence Center, a branch of the U.S. military, is a center dedicated to the procurement and implementation of AI software and neuromorphic hardware for combat use. Specific applications include smart headsets/goggles and robots. JAIC intends to rely heavily on neuromorphic technology to connect "every fighter every shooter" within a network of neuromorphic-enabled units.<br />
<br />
军民两用联合人工智能中心是美国军队的一个分支，专门从事采购和实施用于战斗的人工智能软件和神经形态硬件。具体应用包括智能耳机/护目镜和机器人。JAIC 打算严重依赖神经形态技术来连接神经形态单位网络中的“每个战士每个射手”。<br />
<br />
==Legal considerations==<br />
==法律问题==<br />
Skeptics have argued that there is no way to apply the electronic personhood, the concept of personhood that would apply to neuromorphic technology, legally. In a letter signed by 285 experts in law, robotics, medicine, and ethics opposing a European Commission proposal to recognize “smart robots” as legal persons, the authors write, “A legal status for a robot can’t derive from the [[Natural person|Natural Person]] model, since the robot would then hold [[human rights]], such as the right to dignity, the right to its integrity, the right to remuneration or the right to citizenship, thus directly confronting the Human rights. This would be in contradiction with the [[Charter of Fundamental Rights of the European Union]] and the [[Convention for the Protection of Human Rights and Fundamental Freedoms]].”<ref>{{Cite web|url=http://www.robotics-openletter.eu/|title=Robotics Openletter {{!}} Open letter to the European Commission|language=fr-FR|access-date=2019-05-10}}</ref><br />
<br />
Skeptics have argued that there is no way to apply the electronic personhood, the concept of personhood that would apply to neuromorphic technology, legally. In a letter signed by 285 experts in law, robotics, medicine, and ethics opposing a European Commission proposal to recognize “smart robots” as legal persons, the authors write, “A legal status for a robot can’t derive from the Natural Person model, since the robot would then hold human rights, such as the right to dignity, the right to its integrity, the right to remuneration or the right to citizenship, thus directly confronting the Human rights. This would be in contradiction with the Charter of Fundamental Rights of the European Union and the Convention for the Protection of Human Rights and Fundamental Freedoms.”<br />
<br />
怀疑论者认为，在法律上没有办法应用电子人格，这个人格概念将适用于神经形态技术。在一封由285名法律、机器人、医学和伦理学专家签名的信中，作者们反对欧盟委员会承认“智能机器人”为法人的提议，他们写道，“机器人的法律地位不能从自然人模型中推导出来，因为机器人将拥有人权，如尊严权、完整权、报酬权或公民权，从而直接面对人权。这将有悖于《欧洲联盟基本权利宪章和《保护人权和基本自由公约》。”<br />
<br />
===Ownership and property rights===<br />
===所有权及财产权问题===<br />
There is significant legal debate around property rights and artificial intelligence. In ''Acohs Pty Ltd v. Ucorp Pty Ltd'', Justice Christopher Jessup of the [[Federal Court of Australia]] found that the [[source code]] for [[Material safety data sheets|Material Safety Data Sheets]] could not be [[Copyright law of Australia|copyrighted]] as it was generated by a [[software interface]] rather than a human author.<ref>{{Cite web|url=http://www.lavan.com.au/advice/intellectual_property_technology/copyright_in_source_code_and_digital_products|title=Copyright in source code and digital products|last=Lavan|website=Lavan|language=en|access-date=2019-05-10}}</ref> The same question may apply to neuromorphic systems: if a neuromorphic system successfully mimics a human brain and produces a piece of original work, who, if anyone, should be able to claim ownership of the work?<ref>{{cite journal |last1=Eshraghian|first1=Jason K. |title=Human Ownership of Artificial Creativity |journal=Nature Machine Intelligence |date=9 March 2020 |volume=2 |pages=157–160 |doi=10.1038/s42256-020-0161-x}}</ref><br />
<br />
There is significant legal debate around property rights and artificial intelligence. In Acohs Pty Ltd v. Ucorp Pty Ltd, Justice Christopher Jessup of the Federal Court of Australia found that the source code for Material Safety Data Sheets could not be copyrighted as it was generated by a software interface rather than a human author. The same question may apply to neuromorphic systems: if a neuromorphic system successfully mimics a human brain and produces a piece of original work, who, if anyone, should be able to claim ownership of the work?<br />
<br />
围绕财产权和人工智能有着重大的法律争论。在 Acohs Pty Ltd 诉 Ucorp Pty Ltd 一案中，澳大利亚联邦法院的克里斯托弗 · 杰瑟普法官发现，材料安全数据表的源代码不能受版权保护，因为它是由软件界面而不是人工作者生成的。同样的问题可能也适用于神经形态系统: 如果一个神经形态系统成功地模仿了人类的大脑并产生了一部原创作品，那么谁，如果有人，应该声称拥有这部作品的所有权？<br />
<br />
==Neuromemristive systems==<br />
==神经记忆电阻系统==<br />
Neuromemristive systems are a subclass of neuromorphic computing systems that focus on the use of [[memristors]] to implement [[neuroplasticity]]. While neuromorphic engineering focuses on mimicking biological behavior, neuromemristive systems focus on abstraction.<ref>{{Cite web|url=https://digitalops.sandia.gov/Mediasite/Play/a10cf6ceb55d47608bb8326dd00e46611d|title=002.08 N.I.C.E. Workshop 2014: Towards Intelligent Computing with Neuromemristive Circuits and Systems - Feb. 2014|website=digitalops.sandia.gov|access-date=2019-08-26}}</ref> For example, a neuromemristive system may replace the details of a [[Cerebral cortex|cortical]] microcircuit's behavior with an abstract neural network model.<ref>C. Merkel and D. Kudithipudi, "Neuromemristive extreme learning machines for pattern classification," ISVLSI, 2014.</ref><br />
<br />
Neuromemristive systems are a subclass of neuromorphic computing systems that focus on the use of memristors to implement neuroplasticity. While neuromorphic engineering focuses on mimicking biological behavior, neuromemristive systems focus on abstraction. For example, a neuromemristive system may replace the details of a cortical microcircuit's behavior with an abstract neural network model.C. Merkel and D. Kudithipudi, "Neuromemristive extreme learning machines for pattern classification," ISVLSI, 2014.<br />
<br />
神经记忆电阻系统是神经形态计算系统的一个亚类，主要研究利用记忆电阻器实现神经可塑性。神经形态工程的重点是模拟生物行为，而神经记忆电阻系统的重点是提取。例如，一个神经记忆系统可以用一个抽象的神经网络模型取代皮层微电路的行为细节。默克尔和 d. Kudithipudi，“用于模式分类的神经记忆极端学习机器”，ISVLSI，2014。<br />
<br />
There exist several neuron inspired threshold logic functions<ref name="Maan 1–13" /> implemented with memristors that have applications in high level [[pattern recognition]] applications. Some of the applications reported recently include [[speech recognition]],<ref>{{Cite journal|title = Memristor pattern recogniser: isolated speech word recognition|journal = Electronics Letters|pages = 1370–1372|volume = 51|issue = 17|doi = 10.1049/el.2015.1428|first1 = A.K.|last1 = Maan|first2 = A.P.|last2 = James|first3 = S.|last3 = Dimitrijev|year = 2015|bibcode = 2015ElL....51.1370M|hdl = 10072/140989|s2cid = 61454815|url = https://semanticscholar.org/paper/48d3ab11ec6e213b62f11eedcfb7b7febb058674|hdl-access = free}}</ref> [[face recognition]]<ref>{{Cite journal|title = Memristive Threshold Logic Face Recognition|journal = Procedia Computer Science|date = 2014-01-01|pages = 98–103|volume = 41|series = 5th Annual International Conference on Biologically Inspired Cognitive Architectures, 2014 BICA|doi = 10.1016/j.procs.2014.11.090|first1 = Akshay Kumar|last1 = Maan|first2 = Dinesh S.|last2 = Kumar|first3 = Alex Pappachen|last3 = James|doi-access = free}}</ref> and [[object recognition]].<ref>{{Cite journal|title = Memristive Threshold Logic Circuit Design of Fast Moving Object Detection|journal = IEEE Transactions on Very Large Scale Integration (VLSI) Systems|date = 2015-10-01|issn = 1063-8210|pages = 2337–2341|volume = 23|issue = 10|doi = 10.1109/TVLSI.2014.2359801|first1 = A.K.|last1 = Maan|first2 = D.S.|last2 = Kumar|first3 = S.|last3 = Sugathan|first4 = A.P.|last4 = James|arxiv = 1410.1267|s2cid = 9647290}}</ref> They also find applications in replacing conventional digital logic gates.<ref>{{Cite journal|title = Resistive Threshold Logic|journal = IEEE Transactions on Very Large Scale Integration (VLSI) Systems|date = 2014-01-01|issn = 1063-8210|pages = 190–195|volume = 22|issue = 1|doi = 10.1109/TVLSI.2012.2232946|first1 = A.P.|last1 = James|first2 = L.R.V.J.|last2 = Francis|first3 = D.S.|last3 = Kumar|arxiv = 1308.0090|s2cid = 7357110}}</ref><ref>{{Cite journal|title = Threshold Logic Computing: Memristive-CMOS Circuits for Fast Fourier Transform and Vedic Multiplication|journal = IEEE Transactions on Very Large Scale Integration (VLSI) Systems|date = 2015-11-01|issn = 1063-8210|pages = 2690–2694|volume = 23|issue = 11|doi = 10.1109/TVLSI.2014.2371857|first1 = A.P.|last1 = James|first2 = D.S.|last2 = Kumar|first3 = A.|last3 = Ajayan|arxiv = 1411.5255|s2cid = 6076956}}</ref><br />
<br />
There exist several neuron inspired threshold logic functions implemented with memristors that have applications in high level pattern recognition applications. Some of the applications reported recently include speech recognition, face recognition and object recognition. They also find applications in replacing conventional digital logic gates.<br />
<br />
利用记忆电阻器实现的神经元启发阈值逻辑函数在高级模式识别中有着广泛的应用。最近报道的一些应用包括语音识别、人脸识别和物体识别。它们还可以用来取代传统的数字逻辑门。<br />
<br />
For ideal passive memristive circuits there is an exact equation (Caravelli-Traversa-[[Di Ventra]] equation) for the internal memory of the circuit:<ref>{{cite journal |last=Caravelli |display-authors=etal|arxiv=1608.08651 |title=The complex dynamics of memristive circuits: analytical results and universal slow relaxation |year=2017 |doi=10.1103/PhysRevE.95.022140 |pmid= 28297937 |volume=95 |issue= 2 |pages= 022140 |journal=Physical Review E|bibcode=2017PhRvE..95b2140C |s2cid=6758362}}</ref><br />
<br />
For ideal passive memristive circuits there is an exact equation (Caravelli-Traversa-Di Ventra equation) for the internal memory of the circuit:<br />
<br />
对于理想的无源记忆电路，电路的内部记忆有一个精确的方程(Caravelli-Traversa-Di Ventra 方程) :<br />
<br />
:<math> \frac{d}{dt} \vec{W} = \alpha \vec{W}-\frac{1}{\beta} (I+\xi \Omega W)^{-1} \Omega \vec S </math><br />
<br />
:\frac{d}{dt} \vec{W} = \alpha \vec{W}-\frac{1}{\beta} (I+\xi \Omega W)^{-1} \Omega \vec S<br />
<br />
:\frac{d}{dt} \vec{W} = \alpha \vec{W}-\frac{1}{\beta} (I+\xi \Omega W)^{-1} \Omega \vec S<br />
<br />
as a function of the properties of the physical memristive network and the external sources. In the equation above, <math>\alpha</math> is the "forgetting" time scale constant, <math> \xi=r-1</math> and <math>r=\frac{R_\text{off}}{R_\text{on}}</math> is the ratio of ''off'' and ''on'' values of the limit resistances of the memristors, <math> \vec S </math> is the vector of the sources of the circuit and <math>\Omega</math> is a projector on the fundamental loops of the circuit. The constant <math>\beta</math> has the dimension of a voltage and is associated to the properties of the [[memristor]]; its physical origin is the charge mobility in the conductor. The diagonal matrix and vector <math>W=\operatorname{diag}(\vec W)</math> and <math>\vec W</math> respectively, are instead the internal value of the memristors, with values between 0 and 1. This equation thus requires adding extra constraints on the memory values in order to be reliable.<br />
<br />
<nowiki>as a function of the properties of the physical memristive network and the external sources. In the equation above, \alpha is the "forgetting" time scale constant, \xi=r-1 and r=\frac{R_\text{off}}{R_\text{on}} is the ratio of off and on values of the limit resistances of the memristors, \vec S is the vector of the sources of the circuit and \Omega is a projector on the fundamental loops of the circuit. The constant \beta has the dimension of a voltage and is associated to the properties of the memristor; its physical origin is the charge mobility in the conductor. The diagonal matrix and vector W=\operatorname{diag}(\vec W) and \vec W respectively, are instead the internal value of the memristors, with values between 0 and 1. This equation thus requires adding extra constraints on the memory values in order to be reliable.</nowiki><br />
<br />
<nowiki>作为物理记忆网络和外部源的性质的函数。在上述方程中，α 是“遗忘”时间尺度常数，xi = r-1，r = frac { r _ text { off }{ on }{ r _ text { on }}是记忆电阻器极限电阻的开关和开关值之比，vec s 是电路源的矢量，Omega 是电路基本环路的投影仪。常数 β 具有电压的尺寸，与记忆电阻器的特性有关; 它的物理起源是导体中的电荷迁移率。对角矩阵和向量 w = 操作者名{ diag }(vec w)和 vec w 分别是记忆电阻器的内值，值在0到1之间。因此，这个等式需要在内存值上添加额外的约束，以保证可靠性。</nowiki><br />
<br />
==See also==<br />
== 相关词条 ==<br />
{{Columns-list|colwidth=18em|<br />
* [[AI accelerator (computer hardware)|AI accelerator]]<br />
* [[Artificial brain]]<br />
* [[Biomorphic]]<br />
* [[Cognitive computer]]<br />
* [[Computation and Neural Systems]]<br />
* [[Differentiable programming]]<br />
* [[Event camera]]<br />
* [[Neurorobotics]]<br />
* [[Optical flow sensor]]<br />
* [[Physical neural network]]<br />
* [[SpiNNaker]]<br />
* [[SyNAPSE]]<br />
* [[Retinomorphic sensor]]<br />
* [[Vision chip]]<br />
* [[Vision processing unit]]<br />
* [[Zeroth (software)]]<br />
* [[Hardware for artificial intelligence]]<br />
}}<br />
<br />
{{Portal bar|Electronics}}<br />
<br />
==References==<br />
{{Reflist|40em}}<br />
<br />
==External links==<br />
<!--======================== {{No more links}} ============================<br />
| PLEASE BE CAUTIOUS IN ADDING MORE LINKS TO THIS ARTICLE. Wikipedia |<br />
| is not a collection of links nor should it be used for advertising. |<br />
| |<br />
| Excessive or inappropriate links WILL BE DELETED. |<br />
| See [[Wikipedia:External links]] & [[Wikipedia:Spam]] for details. |<br />
| |<br />
| If there are already plentiful links, please propose additions or |<br />
| replacements on this article's discussion page, or submit your link |<br />
| to the relevant category at the Open Directory Project (dmoz.org) |<br />
| and link back to that category using the {{dmoz}} template. |<br />
======================={{No more links}}=============================--><br />
{{Commons category|Neuromorphic Engineering}}<br />
*[https://web.archive.org/web/20150727034331/http://ine-web.org/workshops/workshops-overview Telluride Neuromorphic Engineering Workshop]<br />
*[https://archive.today/20130115190057/http://capocaccia.ethz.ch/ CapoCaccia Cognitive Neuromorphic Engineering Workshop]<br />
*[https://web.archive.org/web/20190716132350/http://www.ine-web.org/ Institute of Neuromorphic Engineering]<br />
*[http://www.ine-news.org/ INE news site].<br />
*[http://www.frontiersin.org/neuromorphic_engineering Frontiers in Neuromorphic Engineering Journal]<br />
*[http://www.cns.caltech.edu/ Computation and Neural Systems] department at the [[California Institute of Technology]].<br />
*[http://www.humanbrainproject.eu/ Human Brain Project official site]<br />
*[https://www.the-scientist.com/features/building-a-silicon-brain-65738 Building a Silicon Brain:] Computer chips based on biological neurons may help simulate larger and more-complex brain models. May 1, 2019. SANDEEP RAVINDRAN<br />
<br />
<br />
*Telluride Neuromorphic Engineering Workshop<br />
*CapoCaccia Cognitive Neuromorphic Engineering Workshop<br />
* Institute of Neuromorphic Engineering<br />
*INE news site.<br />
*Frontiers in Neuromorphic Engineering Journal<br />
*Computation and Neural Systems department at the California Institute of Technology.<br />
*Human Brain Project official site<br />
*Building a Silicon Brain: Computer chips based on biological neurons may help simulate larger and more-complex brain models. May 1, 2019. SANDEEP RAVINDRAN<br />
<br />
=<nowiki>外部链接</nowiki>= <br />
*碲化物神经形态工程工作室<br />
*CapoCaccia 认知神经形态工程工作室<br />
*神经形态工程研究所<br />
*INE 新闻站点。<br />
*《神经形态工程学前沿》<br />
*加州理工学院计算与神经系统系。<br />
*人脑项目官方网站<br />
*打造硅脑: 基于生物神经元的计算机芯片可能有助于模拟更大、更复杂的大脑模型。2019年5月1日。SANDEEP RAVINDRAN<br />
<br />
{{Differentiable computing}}<br />
{{Authority control}}<br />
<br />
[[Category:Electrical engineering]]<br />
[[Category:Neuroscience]]<br />
[[分类:AI accelerators|*]]<br />
[[Category:Artificial intelligence]]<br />
[[Category:Robotics]]<br />
<br />
Category:Electrical engineering<br />
Category:Neuroscience<br />
*<br />
Category:Artificial intelligence<br />
Category:Robotics<br />
<br />
类别: 电气工程类别: 神经科学 * 类别: 人工智能类别: 机器人<br />
<br />
<noinclude><br />
<br />
<small>This page was moved from [[wikipedia:en:Neuromorphic engineering]]. Its edit history can be viewed at [[神经形态计算/edithistory]]</small></noinclude><br />
<br />
[[Category:待整理页面]]<br />
<br />
<references /></div>Shenky20https://wiki.swarma.org/index.php?title=%E7%A5%9E%E7%BB%8F%E5%BD%A2%E6%80%81%E8%AE%A1%E7%AE%97&diff=30178神经形态计算2022-04-13T08:28:40Z<p>Shenky20：/* 神经生物学启发 */</p>
<hr />
<div><br />
此词条由神经动力学模型读书会词条梳理志愿者Shenky20翻译审校，未经专家审校，带来阅读不便，请见谅。<br />
<br />
{{Use American English|date = January 2019}}<br />
{{Use mdy dates|date = January 2019}}<br />
<br />
'''Neuromorphic engineering''', also known as '''neuromorphic computing''',<ref name=":3">{{Cite journal | doi = 10.1145/2601069| title = Neuromorphic computing gets ready for the (really) big time| journal = [[Communications of the ACM]]| volume = 57| issue = 6| year = 2014| last1 = Monroe | first1 = D. | pages = 13–15| s2cid = 20051102}}</ref><ref name=":4">{{Cite journal | doi = 10.1088/0957-4484/21/17/175202| pmid = 20368686| title = Nanotube devices based crossbar architecture: Toward neuromorphic computing| journal = Nanotechnology| volume = 21| issue = 17| pages = 175202| year = 2010| last1 = Zhao | first1 = W. S. | last2 = Agnus | first2 = G. | last3 = Derycke | first3 = V. | last4 = Filoramo | first4 = A. | last5 = Bourgoin | first5 = J. -P. | last6 = Gamrat | first6 = C. | bibcode = 2010Nanot..21q5202Z| url = https://zenodo.org/record/3428659}}</ref><ref name="humanbrainproject">{{YouTube|id=6RoiZ90mGfw|title=The Human Brain Project SP 9: Neuromorphic Computing Platform}}</ref> is the use of [[very-large-scale integration]] (VLSI) systems containing electronic [[analog circuit]]s to mimic neuro-biological architectures present in the nervous system. A neuromorphic computer/chip is any device that uses physical artificial neurons (made from silicon) to do computations.<ref name=":5">{{cite journal|last1=Mead|first1=Carver|title=Neuromorphic electronic systems|journal=Proceedings of the IEEE|date=1990|volume=78|issue=10|pages=1629–1636|doi=10.1109/5.58356|url=https://authors.library.caltech.edu/53090/1/00058356.pdf}}</ref><ref name=":2" /> In recent times, the term ''neuromorphic'' has been used to describe analog, digital, mixed-mode analog/digital VLSI, and software systems that implement models of [[neural system]]s (for [[perception]], [[motor control]], or [[multisensory integration]]). The implementation of neuromorphic computing on the hardware level can be realized by oxide-based [[memristor]]s,<ref name="Maan 1–13">{{Cite journal|last1=Maan|first1=A. K.|last2=Jayadevi|first2=D. A.|last3=James|first3=A. P.|date=2016-01-01|title=A Survey of Memristive Threshold Logic Circuits|journal=IEEE Transactions on Neural Networks and Learning Systems|volume=PP|issue=99|pages=1734–1746|doi=10.1109/TNNLS.2016.2547842|pmid=27164608|issn=2162-237X|arxiv=1604.07121|bibcode=2016arXiv160407121M|s2cid=1798273}}</ref> spintronic memories, threshold switches, and [[transistor]]s.<ref name=":6">{{Cite journal|title = Mott Memory and Neuromorphic Devices|journal = Proceedings of the IEEE|date = 2015-08-01|issn = 0018-9219|pages = 1289–1310|volume = 103|issue = 8|doi = 10.1109/JPROC.2015.2431914|first1 = You|last1 = Zhou|first2 = S.|last2 = Ramanathan|s2cid = 11347598|url = https://zenodo.org/record/895565}}</ref><ref name=":2">{{Cite document|title=Neuromorphic Circuits With Neural Modulation Enhancing the Information Content of Neural Signaling {{!}} International Conference on Neuromorphic Systems 2020|language=EN|doi=10.1145/3407197.3407204|s2cid=220794387}}</ref> Training software-based neuromorphic systems of spiking neural networks can be achieved using error backpropagation, e.g., using [[Python (programming language)|Python]] based frameworks such as snnTorch,<ref name=":7">{{cite journal |last1=Eshraghian|first1=Jason K.|last2=Ward|first2=Max|last3=Neftci |first3=Emre|last4=Wang|first4=Xinxin|last5=Lenz|first5=Gregor|last6=Dwivedi|first6=Girish|last7=Bennamoun|first7=Mohammed|last8=Jeong|first8=Doo Seok|last9=Lu|first9=Wei D.|title=Training Spiking Neural Networks Using Lessons from Deep Learning |date=1 October 2021 |arxiv=2109.12894 }}</ref> or using canonical learning rules from the biological learning literature, e.g., using BindsNet.<ref name=":8">{{Cite web | url=https://github.com/Hananel-Hazan/bindsnet | title=Hananel-Hazan/bindsnet: Simulation of spiking neural networks (SNNs) using PyTorch.| date=31 March 2020}}</ref><br />
<br />
<br />
'''<font color="#ff8000">神经形态工程Neuromorphic engineering</font>'''(也称为'''<font color="#ff8000">神经形态计算Neuromorphic computing</font>''')<ref name=":3" /><ref name=":4" /><ref name="humanbrainproject" />是指使用包含电子'''<font color="#ff8000">模拟电路Analog circuit</font>'''的'''<font color="#ff8000">超大规模集成电路Very-large-scale integration</font>'''系统来模拟神经系统中生理结构的研究方法。神经形态计算机或神经形态芯片包括任何使用由硅制成的人造神经元进行计算的设备。<ref name=":5" /><ref name=":2" />近年来，神经形态学（neuromorphic）这个术语被用来描述能够实现'''<font color="#ff8000">神经系统Neural system</font>'''模型功能(如'''<font color="#ff8000">感知Perception</font>'''、'''<font color="#ff8000">运动控制Motor control</font>'''，'''<font color="#ff8000">多感官整合Multisensory integration</font>'''等)的模拟、数字、模拟/数字混合模式超大规模集成电路和软件系统。神经形态计算的硬件实现可以通过基于氧化物的'''<font color="#ff8000">记忆电阻器Memristor</font>'''、<ref name="Maan 1–13" />自旋电子存储器、阈值开关和'''<font color="#ff8000">晶体管Transistor</font>'''来实现。<ref name=":6" /><ref name=":2" />对基于软件的脉冲神经网络系统的训练可以通过误差反向传播机制来实现，例如，使用snnTorch等基于Python的框架，<ref name=":7" />或使用BindsNet等典型的受生物启发的学习模式。<ref name=":8" /><br />
<br />
A key aspect of neuromorphic engineering is understanding how the morphology of individual neurons, circuits, applications, and overall architectures creates desirable computations, affects how information is represented, influences robustness to damage, incorporates learning and development, adapts to local change (plasticity), and facilitates evolutionary change.<br />
<br />
<br />
神经形态工程领域的一个关键问题，就是理解'''<font color="#32CD32">单个神经元形态、神经回路、应用和整体结构</font>'''如何产生理想的计算，如何影响信息的表示和对破坏的鲁棒性，如何整合学习和发展，如何适应局部变化(可塑性) 并促进逐渐发展的变化。<br />
<br />
Neuromorphic engineering is an interdisciplinary subject that takes inspiration from [[biology]], [[physics]], [[mathematics]], [[computer science]], and [[electronic engineering]]<ref name=":2" /> to design artificial neural systems, such as [[Machine vision|vision systems]], head-eye systems, auditory processors, and autonomous robots, whose physical architecture and design principles are based on those of biological nervous systems.<ref name=":9">{{Cite journal | doi = 10.1155/2012/705483| title = Qualitative Functional Decomposition Analysis of Evolved Neuromorphic Flight Controllers| journal = Applied Computational Intelligence and Soft Computing| volume = 2012| pages = 1–21| year = 2012| last1 = Boddhu | first1 = S. K. | last2 = Gallagher | first2 = J. C. | doi-access = free}}</ref> It was developed by [[Carver Mead]]<ref name=":10">{{cite web|title=carver mead website|last1=Mead|first1=Carver|url=http://www.carvermead.caltech.edu/index.html|website=carvermead}}</ref> in the late 1980s.<br />
<br />
<br />
神经形态工程是以'''<font color="#ff8000">生物学Biology</font>'''、'''<font color="#ff8000">物理学Physics</font>'''、'''<font color="#ff8000">数学Mathematics</font>'''、'''<font color="#ff8000">计算机科学Computer science</font>'''和'''<font color="#ff8000">电子工程Electronic engineering<ref name=":2" /></font>'''等学科为基础，设计人工神经系统（如'''<font color="#ff8000">视觉系统Vision systems</font>'''、头眼系统、听觉处理器以及物理结构和设计原则都受启发于生物神经系统的自主机器人）的一门交叉学科。<ref name=":9" />20世纪80年代后期，卡弗·米德极大地推动了神经形态工程领域的发展。<ref name=":10" /><br />
<br />
==Neurological inspiration==<br />
==神经生物学启发==<br />
Neuromorphic engineering is set apart by the inspiration it takes from what we know about the structure and operations of the [[brain]]. Neuromorphic engineering translates what we know about the brain's function into computer systems. Work has mostly focused on replicating the analog nature of [[biological computation]] and the role of [[neuron]]s in [[cognition]].<br />
<br />
<br />
神经形态工程形成的灵感来源于现已知的'''<font color="#ff8000">大脑Brain</font>'''结构及运作机制知识，它将我们对大脑功能的了解用于研究和优化计算机系统。该领域工作主要集中于对'''<font color="#ff8000">生物计算Biological computation</font>'''模拟特性和'''<font color="#ff8000">神经元Neuron</font>'''在'''<font color="#ff8000">认知Cognition</font>'''中所发挥作用的复刻。<br />
<br />
The biological processes of neurons and their [[synapse]]s are dauntingly complex, and thus very difficult to artificially simulate. A key feature of biological brains is that all of the processing in neurons use analog chemical signals. This makes it hard to replicate brains in computers because the current generation of computers is completely digital. However, the characteristics of these parts can be abstracted into mathematical functions that closely capture the essence of the neuron's operations.<br />
<br />
神经元及其'''<font color="#ff8000">突触Synapse</font>'''的生理学过程极其复杂，难以进行人工模拟。神经元在所有处理过程中都使用化学模拟信号，这是大脑的一个关键生理学特征。这个特征大大增加了在计算机中复制大脑的难度，因为目前的计算机是完全数字化的。然而，部分特征可以抽象为数学函数，这些函数能够紧密捕捉神经元运行的本质。<br />
<br />
The goal of neuromorphic computing is not to perfectly mimic the brain and all of its functions, but instead to extract what is known of its structure and operations to be used in a practical computing system. No neuromorphic system will claim nor attempt to reproduce every element of neurons and synapses, but all adhere to the idea that computation is highly [[distributed processing|distributed]] throughout a series of small computing elements analogous to a neuron. While this sentiment is standard, researchers chase this goal with different methods.<ref name=":11">{{Cite journal | doi = 10.1088/1741-2560/13/5/051001| title = Large-scale neuromorphic computing systems| journal = Journal of Neural Engineering| volume = 13| pages = 1–15| year = 2016| last1 = Furber | first1 = Steve| issue = 5| pmid = 27529195| bibcode = 2016JNEng..13e1001F| doi-access = free}}</ref><br />
<br />
神经形态计算的目标不是完美地模拟大脑及其所有功能，而是利用已知的大脑结构和运转机制来研发或优化实际的计算系统。任何神经形态学系统都不会声称或试图复制神经元和突触中的每一个元素，但所有人都一致认可将计算高度'''<font color="#ff8000">分散Distribute</font>'''于一系列类似于神经元的小型计算元素的理念。研究人员用不同的方法来追求这一普遍目标。<ref name=":11" /><br />
<br />
<br />
==Examples==<br />
==案例==<br />
As early as 2006, researchers at [[Georgia Tech]] published a field programmable neural array.<ref>{{Cite book|title = A field programmable neural array|last1 = Farquhar|first1 = Ethan|date = May 2006|journal = IEEE International Symposium on Circuits and Systems|pages = 4114–4117|last2 = Hasler|first2 = Paul.|doi = 10.1109/ISCAS.2006.1693534|isbn = 978-0-7803-9389-9|s2cid = 206966013}}</ref> This chip was the first in a line of increasingly complex arrays of floating gate transistors that allowed programmability of charge on the gates of [[MOSFET]]s to model the channel-ion characteristics of neurons in the brain and was one of the first cases of a silicon programmable array of neurons.<br />
<br />
As early as 2006, researchers at Georgia Tech published a field programmable neural array. This chip was the first in a line of increasingly complex arrays of floating gate transistors that allowed programmability of charge on the gates of MOSFETs to model the channel-ion characteristics of neurons in the brain and was one of the first cases of a silicon programmable array of neurons.<br />
<br />
早在2006年，佐治亚理工学院的研究人员就发表了现场可编程神经阵列。这种芯片是一系列越来越复杂的浮栅晶体管阵列中的第一个，这些晶体管可以在 mosfet 的栅极上编程来模拟大脑中神经元的通道离子特性，这也是硅可编程神经元阵列的首批案例之一。<br />
<br />
In November 2011, a group of [[MIT]] researchers created a computer chip that mimics the analog, ion-based communication in a synapse between two neurons using 400 transistors and standard [[CMOS]] manufacturing techniques.<ref>{{cite web|title=MIT creates "brain chip"|url=http://www.extremetech.com/extreme/105067-mit-creates-brain-chip|access-date=4 December 2012}}</ref><ref name="Neuromorphic silicon paper">{{cite journal|title=Neuromorphic silicon neurons and large-scale neural networks: challenges and opportunities|doi=10.3389/fnins.2011.00108|pmid=21991244|pmc=3181466|volume=5|pages=108|journal=Frontiers in Neuroscience|year=2011|last1=Poon|first1=Chi-Sang|last2=Zhou|first2=Kuan|doi-access=free}}</ref><br />
<br />
In November 2011, a group of MIT researchers created a computer chip that mimics the analog, ion-based communication in a synapse between two neurons using 400 transistors and standard CMOS manufacturing techniques.<br />
<br />
2011年11月，麻省理工学院的一组研究人员发明了一种计算机芯片，该芯片使用400个晶体管和标准的 CMOS 制造技术，在两个神经元之间的突触中模拟模拟基于离子的通讯。<br />
<br />
In June 2012, [[spintronic]] researchers at [[Purdue University]] presented a paper on the design of a neuromorphic chip using [[Spin valve|lateral spin valve]]s and [[memristor]]s. They argue that the architecture works similarly to neurons and can therefore be used to test methods of reproducing the brain's processing. In addition, these chips are significantly more energy-efficient than conventional ones.<ref name="Spin Devices Prop">{{Cite arXiv|title=Proposal For Neuromorphic Hardware Using Spin Devices|eprint=1206.3227|last1=Sharad|first1=Mrigank|last2=Augustine|first2=Charles|last3=Panagopoulos|first3=Georgios|last4=Roy|first4=Kaushik|class=cond-mat.dis-nn|year=2012}}</ref><br />
<br />
In June 2012, spintronic researchers at Purdue University presented a paper on the design of a neuromorphic chip using lateral spin valves and memristors. They argue that the architecture works similarly to neurons and can therefore be used to test methods of reproducing the brain's processing. In addition, these chips are significantly more energy-efficient than conventional ones.<br />
<br />
2012年6月，普渡大学的自旋电子学研究人员发表了一篇关于利用侧向自旋阀和记忆电阻器设计神经形态芯片的论文。他们认为，这种结构的工作原理与神经元相似，因此可以用来测试复制大脑处理过程的方法。此外，这些芯片明显比传统芯片更节能。<br />
<br />
Research at [[HP Labs]] on Mott memristors has shown that while they can be non-[[Volatile memory|volatile]], the volatile behavior exhibited at temperatures significantly below the [[phase transition]] temperature can be exploited to fabricate a [[neuristor]],<ref name=":0" /> a biologically-inspired device that mimics behavior found in neurons.<ref name=":0">{{Cite journal | doi = 10.1038/nmat3510| pmid = 23241533| title = A scalable neuristor built with Mott memristors| journal = Nature Materials| volume = 12| issue = 2| pages = 114–7| year = 2012| last1 = Pickett | first1 = M. D. | last2 = Medeiros-Ribeiro | first2 = G. | last3 = Williams | first3 = R. S. | bibcode = 2013NatMa..12..114P| s2cid = 16271627| url = https://semanticscholar.org/paper/b6ba6f496ace2b947f111059663e76bb60e9efeb}}</ref> In September 2013, they presented models and simulations that show how the spiking behavior of these neuristors can be used to form the components required for a [[Turing machine]].<ref>{{cite journal|doi=10.1088/0957-4484/24/38/384002|title=Phase transitions enable computational universality in neuristor-based cellular automata|author1=Matthew D Pickett|author2=R Stanley Williams|name-list-style=amp|date=September 2013|publisher=IOP Publishing Ltd|journal=Nanotechnology|volume=24|issue=38|pmid=23999059|bibcode=2013Nanot..24L4002P|at=384002}}</ref><br />
<br />
Research at HP Labs on Mott memristors has shown that while they can be non-volatile, the volatile behavior exhibited at temperatures significantly below the phase transition temperature can be exploited to fabricate a neuristor, a biologically-inspired device that mimics behavior found in neurons. In September 2013, they presented models and simulations that show how the spiking behavior of these neuristors can be used to form the components required for a Turing machine.<br />
<br />
惠普实验室在莫特记忆电阻器上的研究表明，尽管它们可以是非挥发性的，但是在相变温度以下显示的挥发性行为可以被用来制造神经元电阻器，这是一种模仿神经元行为的生物设备。2013年9月，他们展示了模型和仿真，展示了这些神经元的尖峰行为是如何被用来形成图灵机所需的元件的。<br />
<br />
[[Neurogrid]], built by ''Brains in Silicon'' at [[Stanford University]],<ref>{{cite journal|last1=Boahen|first1=Kwabena|title=Neurogrid: A Mixed-Analog-Digital Multichip System for Large-Scale Neural Simulations|journal=Proceedings of the IEEE|date=24 April 2014|volume=102|issue=5|pages=699–716|doi=10.1109/JPROC.2014.2313565|s2cid=17176371}}</ref> is an example of hardware designed using neuromorphic engineering principles. The circuit board is composed of 16 custom-designed chips, referred to as NeuroCores. Each NeuroCore's analog circuitry is designed to emulate neural elements for 65536 neurons, maximizing energy efficiency. The emulated neurons are connected using digital circuitry designed to maximize spiking throughput.<ref>{{cite journal|doi=10.1038/503022a|pmid = 24201264|title = Neuroelectronics: Smart connections|journal = Nature|volume = 503|issue = 7474|pages = 22–4|year = 2013|last1 = Waldrop|first1 = M. Mitchell|bibcode = 2013Natur.503...22W|doi-access = free}}</ref><ref>{{cite journal|doi=10.1109/JPROC.2014.2313565|title = Neurogrid: A Mixed-Analog-Digital Multichip System for Large-Scale Neural Simulations|journal = Proceedings of the IEEE|volume = 102|issue = 5|pages = 699–716|year = 2014|last1 = Benjamin|first1 = Ben Varkey|last2 = Peiran Gao|last3 = McQuinn|first3 = Emmett|last4 = Choudhary|first4 = Swadesh|last5 = Chandrasekaran|first5 = Anand R.|last6 = Bussat|first6 = Jean-Marie|last7 = Alvarez-Icaza|first7 = Rodrigo|last8 = Arthur|first8 = John V.|last9 = Merolla|first9 = Paul A.|last10 = Boahen|first10 = Kwabena|s2cid = 17176371}}</ref><br />
<br />
Neurogrid, built by Brains in Silicon at Stanford University, is an example of hardware designed using neuromorphic engineering principles. The circuit board is composed of 16 custom-designed chips, referred to as NeuroCores. Each NeuroCore's analog circuitry is designed to emulate neural elements for 65536 neurons, maximizing energy efficiency. The emulated neurons are connected using digital circuitry designed to maximize spiking throughput.<br />
<br />
神经网格是由斯坦福大学硅谷大脑公司建立的，是一个使用神经形态工程原理设计硬件的例子。该电路板由16个定制设计的芯片组成，称为 NeuroCores。每个 NeuroCore 的模拟电路被设计为模拟65536个神经元的神经元元件，最大限度地提高能量效率。模拟的神经元通过设计的数字电路连接，以最大化脉冲吞吐量。<br />
<br />
A research project with implications for neuromorphic engineering is the [[Human Brain Project]] that is attempting to simulate a complete human brain in a supercomputer using biological data. It is made up of a group of researchers in neuroscience, medicine, and computing.<ref>{{cite web|title=Involved Organizations|url=http://www.humanbrainproject.eu/partners.html|access-date=22 February 2013|url-status=dead|archive-url=https://web.archive.org/web/20130302142627/http://www.humanbrainproject.eu/partners.html|archive-date=2 March 2013}}</ref> [[Henry Markram]], the project's co-director, has stated that the project proposes to establish a foundation to explore and understand the brain and its diseases, and to use that knowledge to build new computing technologies. The three primary goals of the project are to better understand how the pieces of the brain fit and work together, to understand how to objectively diagnose and treat brain diseases, and to use the understanding of the human brain to develop neuromorphic computers. That the simulation of a complete human brain will require a supercomputer a thousand times more powerful than today's encourages the current focus on neuromorphic computers.<ref>{{cite web|title=Human Brain Project|url=http://www.humanbrainproject.eu|access-date=22 February 2013}}</ref> $1.3 billion has been allocated to the project by The [[European Commission]].<ref>{{cite web|title=The Human Brain Project and Recruiting More Cyberwarriors|url=http://www.marketplace.org/topics/tech/human-brain-project-and-recruiting-more-cyberwarriors|access-date=22 February 2013|date=January 29, 2013}}</ref><br />
<br />
A research project with implications for neuromorphic engineering is the Human Brain Project that is attempting to simulate a complete human brain in a supercomputer using biological data. It is made up of a group of researchers in neuroscience, medicine, and computing. Henry Markram, the project's co-director, has stated that the project proposes to establish a foundation to explore and understand the brain and its diseases, and to use that knowledge to build new computing technologies. The three primary goals of the project are to better understand how the pieces of the brain fit and work together, to understand how to objectively diagnose and treat brain diseases, and to use the understanding of the human brain to develop neuromorphic computers. That the simulation of a complete human brain will require a supercomputer a thousand times more powerful than today's encourages the current focus on neuromorphic computers. $1.3 billion has been allocated to the project by The European Commission.<br />
<br />
一个对神经形态工程有影响的研究项目是人脑项目，它试图用生物数据在超级计算机中模拟完整的人脑。它由一群神经科学、医学和计算机科学的研究人员组成。该项目的联合主管亨利•马克拉姆(Henry Markram)表示，该项目提议建立一个基础，以探索和了解大脑及其疾病，并利用这些知识构建新的计算机技术。这个项目的三个主要目标是: 更好地理解大脑的各个部分是如何相互匹配和协同工作的; 理解如何客观地诊断和治疗脑部疾病; 以及利用对人类大脑的理解来开发神经形态计算机。模拟一个完整的人类大脑需要一台比现在强大一千倍的超级计算机，这鼓励了当前对神经形态计算机的关注。欧洲委员会已经拨款13亿美元用于这个项目。<br />
<br />
Other research with implications for neuromorphic engineering involves the [[BRAIN Initiative]]<ref name="economist">[https://www.economist.com/news/science-and-technology/21582495-computers-will-help-people-understand-brains-better-and-understanding-brains Neuromorphic computing: The machine of a new soul], The Economist, 2013-08-03</ref> and the [[TrueNorth]] chip from [[IBM]].<ref>{{cite journal|last1=Modha|first1=Dharmendra|title=A million spiking-neuron integrated circuit with a scalable communication network and interface|journal=Science|date=Aug 2014|volume=345|issue=6197|pages=668–673|doi=10.1126/science.1254642|pmid=25104385|bibcode=2014Sci...345..668M|s2cid=12706847}}</ref> Neuromorphic devices have also been demonstrated using nanocrystals, nanowires, and conducting polymers.<ref>{{Cite web|url=http://jessamynfairfield.com/wp-content/uploads/2017/03/PWMar17Fairfield.pdf|title=Smarter Machines|last=Fairfield|first=Jessamyn|date=March 1, 2017}}</ref><br />
<br />
Other research with implications for neuromorphic engineering involves the BRAIN InitiativeNeuromorphic computing: The machine of a new soul, The Economist, 2013-08-03 and the TrueNorth chip from IBM. Neuromorphic devices have also been demonstrated using nanocrystals, nanowires, and conducting polymers.<br />
<br />
其他与神经形态工程有关的研究还包括 BRAIN initiativity/euromorphic computing: The machine of a new soul，The Economist，2013-08-03 and The TrueNorth chip from IBM。神经形态学设备也已经被证明使用纳米晶体、纳米线和导电聚合物。<br />
<br />
[[Intel]] unveiled its neuromorphic research chip, called “[[Intel Loihi|Loihi]]”, in October 2017. The chip uses an asynchronous [[spiking neural network]] (SNN) to implement adaptive self-modifying event-driven fine-grained parallel computations used to implement learning and inference with high efficiency.<ref>{{cite journal |last1=Davies |first1=Mike |title=Loihi: A Neuromorphic Manycore Processor with On-Chip Learning |journal=IEEE Micro |date=January 16, 2018 |volume=38 |issue=1 |pages=82–99 |display-authors=etal|doi=10.1109/MM.2018.112130359 |s2cid=3608458 }}</ref><ref name="Morris2017">{{cite web |last1=Morris |first1=John |title=Why Intel built a neuromorphic chip |url=https://www.zdnet.com/article/why-intel-built-a-neuromorphic-chip/ |website=ZDNet |access-date=17 August 2018 |language=en}}</ref><br />
<br />
Intel unveiled its neuromorphic research chip, called “Loihi”, in October 2017. The chip uses an asynchronous spiking neural network (SNN) to implement adaptive self-modifying event-driven fine-grained parallel computations used to implement learning and inference with high efficiency.<br />
<br />
2017年10月，英特尔公布了其神经形态研究芯片“ Loihi”。该芯片采用异步脉冲神经网络(SNN)实现自适应自修改事件驱动的细粒度并行计算，实现了高效的学习和推理。<br />
<br />
[[IMEC]], a Belgium-based nanoelectronics research center, demonstrated the world's first self-learning neuromorphic chip. The brain-inspired chip, based on OxRAM technology, has the capability of self-learning and has been demonstrated to have the ability to compose music.<ref>{{cite web |title=Imec demonstrates self-learning neuromorphic chip that composes music |url=https://www.imec-int.com/en/articles/imec-demonstrates-self-learning-neuromorphic-chip-that-composes-music |website=IMEC International |access-date=1 October 2019}}</ref> IMEC released the 30-second tune composed by the prototype. The chip was sequentially loaded with songs in the same time signature and style. The songs were old Belgian and French flute minuets, from which the chip learned the rules at play and then applied them.<ref>{{cite web|last1=Bourzac|first1=Katherine|title=A Neuromorphic Chip That Makes Music|url=https://spectrum.ieee.org/a-neuromorphic-chip-that-makes-music|url-status=live|access-date=1 October 2019|website=IEEE Spectrum|date=May 23, 2017}}</ref><br />
<br />
IMEC, a Belgium-based nanoelectronics research center, demonstrated the world's first self-learning neuromorphic chip. The brain-inspired chip, based on OxRAM technology, has the capability of self-learning and has been demonstrated to have the ability to compose music. IMEC released the 30-second tune composed by the prototype. The chip was sequentially loaded with songs in the same time signature and style. The songs were old Belgian and French flute minuets, from which the chip learned the rules at play and then applied them.<br />
<br />
比利时的一个纳米电子学研究中心 IMEC 展示了世界上第一个自学习神经形态芯片。这种基于 OxRAM 技术的大脑启发芯片具有自学习能力，并已被证明具有创作音乐的能力。IMEC 发布了由原型机谱写的30秒曲调。芯片按顺序加载同时签名和风格的歌曲。歌曲是古老的比利时和法国长笛小步舞曲，筹码从中学习游戏规则，然后应用它们。<br />
<br />
[[Blue Brain Project|The Blue Brain Project]], led by Henry Markram, aims to build biologically detailed digital reconstructions and simulations of the mouse brain. The Blue Brain Project has created in silico models of rodent brains, while attempting to replicate as many details about its biology as possible. The supercomputer-based simulations offer new perspectives on understanding the structure and functions of the brain.<br />
<br />
The Blue Brain Project, led by Henry Markram, aims to build biologically detailed digital reconstructions and simulations of the mouse brain. The Blue Brain Project has created in silico models of rodent brains, while attempting to replicate as many details about its biology as possible. The supercomputer-based simulations offer new perspectives on understanding the structure and functions of the brain.<br />
<br />
由亨利 · 马克拉姆领导的蓝脑计划旨在建立小鼠大脑的生物细节数字重建和模拟。蓝色大脑项目已经在啮齿动物大脑的电子模型中创建，同时试图尽可能多地复制其生物学细节。基于超级计算机的模拟为理解大脑的结构和功能提供了新的视角。<br />
<br />
The European Union funded a series of projects at the University of Heidelberg, which led to the development of <br />
[[BrainScaleS]] (brain-inspired multiscale computation in neuromorphic hybrid systems), a hybrid analog [[neuromorphic]] supercomputer located at Heidelberg University, Germany. It was developed as part of the [[Human Brain Project]] neuromorphic computing platform and is the complement to the [[SpiNNaker]] supercomputer (which is based on digital technology). The architecture used in BrainScaleS mimics biological neurons and their connections on a physical level; additionally, since the components are made of silicon, these model neurons operate on average 864 times (24 hours of real time is 100 seconds in the machine simulation) that of their biological counterparts.<ref>{{Cite web|date=2016-03-21|title=Beyond von Neumann, Neuromorphic Computing Steadily Advances|url=https://www.hpcwire.com/2016/03/21/lacking-breakthrough-neuromorphic-computing-steadily-advance/|access-date=2021-10-08|website=HPCwire|language=en-US}}</ref><br />
<br />
The European Union funded a series of projects at the University of Heidelberg, which led to the development of <br />
BrainScaleS (brain-inspired multiscale computation in neuromorphic hybrid systems), a hybrid analog neuromorphic supercomputer located at Heidelberg University, Germany. It was developed as part of the Human Brain Project neuromorphic computing platform and is the complement to the SpiNNaker supercomputer (which is based on digital technology). The architecture used in BrainScaleS mimics biological neurons and their connections on a physical level; additionally, since the components are made of silicon, these model neurons operate on average 864 times (24 hours of real time is 100 seconds in the machine simulation) that of their biological counterparts.<br />
<br />
欧盟资助了海德堡大学的一系列项目，这些项目导致了 BrainScaleS (神经形态混合系统中受大脑启发的多尺度计算)的发展，这是一台位于德国海德堡大学的混合模拟神经形态超级计算机。它是作为人脑计划神经形态计算平台的一部分而开发的，是 SpiNNaker 超级计算机(基于数字技术)的补充。大脑尺度中使用的体系结构模拟了生物神经元及其在物理层面上的连接; 此外，由于这些组件是由硅制成的，这些模型神经元平均运行864倍(在机器模拟中，24小时的实时时间是100秒)。<br />
<br />
===Neuromorphic sensors===<br />
===神经形态传感器===<br />
The concept of neuromorphic systems can be extended to sensors (not just to computation). An example of this applied to detecting [[light]] is the [[retinomorphic sensor]] or, when employed in an array, the [[event camera]].<br />
<br />
The concept of neuromorphic systems can be extended to sensors (not just to computation). An example of this applied to detecting light is the retinomorphic sensor or, when employed in an array, the event camera.<br />
<br />
神经形态系统的概念可以扩展到传感器(而不仅仅是计算)。用于检测光线的一个例子是视网膜变形传感器，或者在阵列中使用的事件摄像机。<br />
<br />
== Ethical considerations==<br />
==伦理问题==<br />
While the interdisciplinary concept of neuromorphic engineering is relatively new, many of the same ethical considerations apply to neuromorphic systems as apply to [[human-like machines]] and [[artificial intelligence]] in general. However, the fact that neuromorphic systems are designed to mimic a [[human brain]] gives rise to unique ethical questions surrounding their usage.<br />
<br />
While the interdisciplinary concept of neuromorphic engineering is relatively new, many of the same ethical considerations apply to neuromorphic systems as apply to human-like machines and artificial intelligence in general. However, the fact that neuromorphic systems are designed to mimic a human brain gives rise to unique ethical questions surrounding their usage.<br />
<br />
虽然神经形态工程的跨学科概念相对较新，但许多同样的伦理考虑适用于神经形态系统，就像适用于类人机器和一般人工智能一样。然而，神经形态系统是为了模仿人类大脑而设计的这一事实引起了围绕其使用的独特的伦理问题。<br />
<br />
However, the practical debate is that neuromorphic hardware as well as artificial "neural networks" are immensely simplified models of how the brain operates or processes information at a much lower [[complex system|complexity in terms of size and functional technology]] and a much more regular structure in terms of [[brain connectivity|connectivity]]. Comparing [[neuromorphic chip]]s to the brain is a very crude comparison similar to comparing a plane to a bird just because they both have wings and a tail. The fact is that neural cognitive systems are many orders of magnitude more [[Energy efficiency (physics)|energy-]] and compute-efficient than current state-of-the-art AI and neuromorphic engineering is an attempt to narrow this gap by inspiring from the brain's mechanism just like many engineering designs have [[bioengineering|bio-inspired features]].<br />
<br />
However, the practical debate is that neuromorphic hardware as well as artificial "neural networks" are immensely simplified models of how the brain operates or processes information at a much lower complexity in terms of size and functional technology and a much more regular structure in terms of connectivity. Comparing neuromorphic chips to the brain is a very crude comparison similar to comparing a plane to a bird just because they both have wings and a tail. The fact is that neural cognitive systems are many orders of magnitude more energy- and compute-efficient than current state-of-the-art AI and neuromorphic engineering is an attempt to narrow this gap by inspiring from the brain's mechanism just like many engineering designs have bio-inspired features.<br />
<br />
然而，实际上的争论是，神经形态硬件和人工“神经网络”是大脑如何运作或处理信息的极其简化的模型，在大小和功能技术方面的复杂性要低得多，在连接方面的结构则更加规则。将神经形态芯片与大脑进行比较是一种非常粗糙的比较，类似于仅仅因为一架飞机有翅膀和一条尾巴就将它与一只鸟进行比较。事实上，神经认知系统比当前最先进的人工智能具有更多的能量和计算效率，而神经形态工程是一种通过从大脑机制中激发灵感来缩小这种差距的尝试，就像许多工程设计具有生物启发的特征一样。数量级。<br />
<br />
===Democratic concerns===<br />
===公众担忧===<br />
Significant ethical limitations may be placed on neuromorphic engineering due to public perception.<ref>{{Cite report|url=https://ai100.stanford.edu/sites/g/files/sbiybj9861/f/ai_100_report_0831fnl.pdf|title=Artificial Intelligence and Life in 2030|author=2015 Study Panel|date=September 2016|work=One Hundred Year Study on Artificial Intelligence (AI100)|publisher=Stanford University}}</ref> Special [[Eurobarometer]] 382: Public Attitudes Towards Robots, a survey conducted by the European Commission, found that 60% of [[European Union]] citizens wanted a ban of robots in the care of children, the elderly, or the disabled. Furthermore, 34% were in favor of a ban on robots in education, 27% in healthcare, and 20% in leisure. The European Commission classifies these areas as notably “human.” The report cites increased public concern with robots that are able to mimic or replicate human functions. Neuromorphic engineering, by definition, is designed to replicate the function of the human brain.<ref name=":1">{{Cite web|url=http://ec.europa.eu/commfrontoffice/publicopinion/archives/ebs/ebs_382_en.pdf|title=Special Eurobarometer 382: Public Attitudes Towards Robots|last=European Commission|date=September 2012|website=European Commission}}</ref><br />
<br />
Significant ethical limitations may be placed on neuromorphic engineering due to public perception. Special Eurobarometer 382: Public Attitudes Towards Robots, a survey conducted by the European Commission, found that 60% of European Union citizens wanted a ban of robots in the care of children, the elderly, or the disabled. Furthermore, 34% were in favor of a ban on robots in education, 27% in healthcare, and 20% in leisure. The European Commission classifies these areas as notably “human.” The report cites increased public concern with robots that are able to mimic or replicate human functions. Neuromorphic engineering, by definition, is designed to replicate the function of the human brain.<br />
<br />
由于公众的认知，神经形态工程学可能会受到严重的伦理限制。欧盟委员会进行的一项调查发现，60% 的欧盟公民希望禁止照顾儿童、老人或残疾人的机器人。此外，34% 的人支持禁止机器人用于教育，27% 的人支持医疗保健，20% 的人支持休闲。欧盟委员会将这些地区明显归类为“人类”报告指出，公众越来越关注能够模仿或复制人类功能的机器人。神经形态工程，顾名思义，是为了复制人脑的功能而设计的。<br />
<br />
The democratic concerns surrounding neuromorphic engineering are likely to become even more profound in the future. The European Commission found that EU citizens between the ages of 15 and 24 are more likely to think of robots as human-like (as opposed to instrument-like) than EU citizens over the age of 55. When presented an image of a robot that had been defined as human-like, 75% of EU citizens aged 15–24 said it corresponded with the idea they had of robots while only 57% of EU citizens over the age of 55 responded the same way. The human-like nature of neuromorphic systems, therefore, could place them in the categories of robots many EU citizens would like to see banned in the future.<ref name=":1" /><br />
<br />
The democratic concerns surrounding neuromorphic engineering are likely to become even more profound in the future. The European Commission found that EU citizens between the ages of 15 and 24 are more likely to think of robots as human-like (as opposed to instrument-like) than EU citizens over the age of 55. When presented an image of a robot that had been defined as human-like, 75% of EU citizens aged 15–24 said it corresponded with the idea they had of robots while only 57% of EU citizens over the age of 55 responded the same way. The human-like nature of neuromorphic systems, therefore, could place them in the categories of robots many EU citizens would like to see banned in the future.<br />
<br />
围绕神经形态工程的民主关注可能在未来变得更加深刻。欧盟委员会(European Commission)发现，15至24岁的欧盟公民比55岁以上的欧盟公民更有可能认为机器人像人(而不是像仪器)。当看到一张被定义为“类人”的机器人图片时，年龄在15岁至24岁之间的欧盟公民中有75% 的人表示，这与他们对机器人的想法相符，而55岁以上的欧盟公民中只有57% 的人有同样的反应。因此，类似人类的神经形态系统，可以把它们归入许多欧盟公民希望在未来禁止使用的机器人类别。<br />
<br />
===Personhood===<br />
===人格问题===<br />
As neuromorphic systems have become increasingly advanced, some scholars{{who|date=August 2021}} have advocated for granting [[personhood]] rights to these systems. If the brain is what grants humans their personhood, to what extent does a neuromorphic system have to mimic the human brain to be granted personhood rights? Critics of technology development in the [[Human Brain Project]], which aims to advance brain-inspired computing, have argued that advancement in neuromorphic computing could lead to machine consciousness or personhood.<ref>{{Cite journal|last=Aicardi|first=Christine|date=September 2018|title=Accompanying technology development in the Human Brain Project: From foresight to ethics management|journal=Futures|volume=102|pages=114–124|doi=10.1016/j.futures.2018.01.005|doi-access=free}}</ref> If these systems are to be treated as people, critics argue, then many tasks humans perform using neuromorphic systems, including the act of termination of neuromorphic systems, may be morally impermissible as these acts would violate the autonomy of the neuromorphic systems.<ref>{{Cite journal|last=Lim|first=Daniel|date=2014-06-01|title=Brain simulation and personhood: a concern with the Human Brain Project|journal=Ethics and Information Technology|language=en|volume=16|issue=2|pages=77–89|doi=10.1007/s10676-013-9330-5|s2cid=17415814|issn=1572-8439}}</ref><br />
<br />
As neuromorphic systems have become increasingly advanced, some scholars have advocated for granting personhood rights to these systems. If the brain is what grants humans their personhood, to what extent does a neuromorphic system have to mimic the human brain to be granted personhood rights? Critics of technology development in the Human Brain Project, which aims to advance brain-inspired computing, have argued that advancement in neuromorphic computing could lead to machine consciousness or personhood. If these systems are to be treated as people, critics argue, then many tasks humans perform using neuromorphic systems, including the act of termination of neuromorphic systems, may be morally impermissible as these acts would violate the autonomy of the neuromorphic systems.<br />
<br />
随着神经形态系统的日益发展，一些学者主张赋予这些系统人格权。如果是大脑赋予了人类人格，那么神经形态系统在多大程度上必须模仿人类大脑才能被赋予人格权利？“人脑计划”旨在推进以大脑为灵感的计算机技术发展，该计划的批评者认为，神经形态计算机技术的进步可能导致机器意识或人格的形成。批评家认为，如果这些系统被当作人来对待，那么人类使用神经形态系统执行的许多任务，包括终止神经形态系统的行为，可能在道德上是不允许的，因为这些行为将违反神经形态系统的自主性。<br />
<br />
==Dual use (military applications)==<br />
==军民两用技术==<br />
The [[Joint Artificial Intelligence Center]], a branch of the U.S. military, is a center dedicated to the procurement and implementation of AI software and neuromorphic hardware for combat use. Specific applications include smart headsets/goggles and robots. JAIC intends to rely heavily on neuromorphic technology to connect "every fighter every shooter" within a network of neuromorphic-enabled units.<br />
<br />
The Joint Artificial Intelligence Center, a branch of the U.S. military, is a center dedicated to the procurement and implementation of AI software and neuromorphic hardware for combat use. Specific applications include smart headsets/goggles and robots. JAIC intends to rely heavily on neuromorphic technology to connect "every fighter every shooter" within a network of neuromorphic-enabled units.<br />
<br />
军民两用联合人工智能中心是美国军队的一个分支，专门从事采购和实施用于战斗的人工智能软件和神经形态硬件。具体应用包括智能耳机/护目镜和机器人。JAIC 打算严重依赖神经形态技术来连接神经形态单位网络中的“每个战士每个射手”。<br />
<br />
==Legal considerations==<br />
== 法律问题 ==<br />
Skeptics have argued that there is no way to apply the electronic personhood, the concept of personhood that would apply to neuromorphic technology, legally. In a letter signed by 285 experts in law, robotics, medicine, and ethics opposing a European Commission proposal to recognize “smart robots” as legal persons, the authors write, “A legal status for a robot can’t derive from the [[Natural person|Natural Person]] model, since the robot would then hold [[human rights]], such as the right to dignity, the right to its integrity, the right to remuneration or the right to citizenship, thus directly confronting the Human rights. This would be in contradiction with the [[Charter of Fundamental Rights of the European Union]] and the [[Convention for the Protection of Human Rights and Fundamental Freedoms]].”<ref>{{Cite web|url=http://www.robotics-openletter.eu/|title=Robotics Openletter {{!}} Open letter to the European Commission|language=fr-FR|access-date=2019-05-10}}</ref><br />
<br />
Skeptics have argued that there is no way to apply the electronic personhood, the concept of personhood that would apply to neuromorphic technology, legally. In a letter signed by 285 experts in law, robotics, medicine, and ethics opposing a European Commission proposal to recognize “smart robots” as legal persons, the authors write, “A legal status for a robot can’t derive from the Natural Person model, since the robot would then hold human rights, such as the right to dignity, the right to its integrity, the right to remuneration or the right to citizenship, thus directly confronting the Human rights. This would be in contradiction with the Charter of Fundamental Rights of the European Union and the Convention for the Protection of Human Rights and Fundamental Freedoms.”<br />
<br />
怀疑论者认为，在法律上没有办法应用电子人格，这个人格概念将适用于神经形态技术。在一封由285名法律、机器人、医学和伦理学专家签名的信中，作者们反对欧盟委员会承认“智能机器人”为法人的提议，他们写道，“机器人的法律地位不能从自然人模型中推导出来，因为机器人将拥有人权，如尊严权、完整权、报酬权或公民权，从而直接面对人权。这将有悖于《欧洲联盟基本权利宪章和《保护人权和基本自由公约》。”<br />
<br />
===Ownership and property rights===<br />
===所有权及财产权问题===<br />
There is significant legal debate around property rights and artificial intelligence. In ''Acohs Pty Ltd v. Ucorp Pty Ltd'', Justice Christopher Jessup of the [[Federal Court of Australia]] found that the [[source code]] for [[Material safety data sheets|Material Safety Data Sheets]] could not be [[Copyright law of Australia|copyrighted]] as it was generated by a [[software interface]] rather than a human author.<ref>{{Cite web|url=http://www.lavan.com.au/advice/intellectual_property_technology/copyright_in_source_code_and_digital_products|title=Copyright in source code and digital products|last=Lavan|website=Lavan|language=en|access-date=2019-05-10}}</ref> The same question may apply to neuromorphic systems: if a neuromorphic system successfully mimics a human brain and produces a piece of original work, who, if anyone, should be able to claim ownership of the work?<ref>{{cite journal |last1=Eshraghian|first1=Jason K. |title=Human Ownership of Artificial Creativity |journal=Nature Machine Intelligence |date=9 March 2020 |volume=2 |pages=157–160 |doi=10.1038/s42256-020-0161-x}}</ref><br />
<br />
There is significant legal debate around property rights and artificial intelligence. In Acohs Pty Ltd v. Ucorp Pty Ltd, Justice Christopher Jessup of the Federal Court of Australia found that the source code for Material Safety Data Sheets could not be copyrighted as it was generated by a software interface rather than a human author. The same question may apply to neuromorphic systems: if a neuromorphic system successfully mimics a human brain and produces a piece of original work, who, if anyone, should be able to claim ownership of the work?<br />
<br />
围绕财产权和人工智能有着重大的法律争论。在 Acohs Pty Ltd 诉 Ucorp Pty Ltd 一案中，澳大利亚联邦法院的克里斯托弗 · 杰瑟普法官发现，材料安全数据表的源代码不能受版权保护，因为它是由软件界面而不是人工作者生成的。同样的问题可能也适用于神经形态系统: 如果一个神经形态系统成功地模仿了人类的大脑并产生了一部原创作品，那么谁，如果有人，应该声称拥有这部作品的所有权？<br />
<br />
==Neuromemristive systems==<br />
==神经记忆电阻系统==<br />
Neuromemristive systems are a subclass of neuromorphic computing systems that focus on the use of [[memristors]] to implement [[neuroplasticity]]. While neuromorphic engineering focuses on mimicking biological behavior, neuromemristive systems focus on abstraction.<ref>{{Cite web|url=https://digitalops.sandia.gov/Mediasite/Play/a10cf6ceb55d47608bb8326dd00e46611d|title=002.08 N.I.C.E. Workshop 2014: Towards Intelligent Computing with Neuromemristive Circuits and Systems - Feb. 2014|website=digitalops.sandia.gov|access-date=2019-08-26}}</ref> For example, a neuromemristive system may replace the details of a [[Cerebral cortex|cortical]] microcircuit's behavior with an abstract neural network model.<ref>C. Merkel and D. Kudithipudi, "Neuromemristive extreme learning machines for pattern classification," ISVLSI, 2014.</ref><br />
<br />
Neuromemristive systems are a subclass of neuromorphic computing systems that focus on the use of memristors to implement neuroplasticity. While neuromorphic engineering focuses on mimicking biological behavior, neuromemristive systems focus on abstraction. For example, a neuromemristive system may replace the details of a cortical microcircuit's behavior with an abstract neural network model.C. Merkel and D. Kudithipudi, "Neuromemristive extreme learning machines for pattern classification," ISVLSI, 2014.<br />
<br />
神经记忆电阻系统是神经形态计算系统的一个亚类，主要研究利用记忆电阻器实现神经可塑性。神经形态工程的重点是模拟生物行为，而神经记忆电阻系统的重点是提取。例如，一个神经记忆系统可以用一个抽象的神经网络模型取代皮层微电路的行为细节。默克尔和 d. Kudithipudi，“用于模式分类的神经记忆极端学习机器”，ISVLSI，2014。<br />
<br />
There exist several neuron inspired threshold logic functions<ref name="Maan 1–13" /> implemented with memristors that have applications in high level [[pattern recognition]] applications. Some of the applications reported recently include [[speech recognition]],<ref>{{Cite journal|title = Memristor pattern recogniser: isolated speech word recognition|journal = Electronics Letters|pages = 1370–1372|volume = 51|issue = 17|doi = 10.1049/el.2015.1428|first1 = A.K.|last1 = Maan|first2 = A.P.|last2 = James|first3 = S.|last3 = Dimitrijev|year = 2015|bibcode = 2015ElL....51.1370M|hdl = 10072/140989|s2cid = 61454815|url = https://semanticscholar.org/paper/48d3ab11ec6e213b62f11eedcfb7b7febb058674|hdl-access = free}}</ref> [[face recognition]]<ref>{{Cite journal|title = Memristive Threshold Logic Face Recognition|journal = Procedia Computer Science|date = 2014-01-01|pages = 98–103|volume = 41|series = 5th Annual International Conference on Biologically Inspired Cognitive Architectures, 2014 BICA|doi = 10.1016/j.procs.2014.11.090|first1 = Akshay Kumar|last1 = Maan|first2 = Dinesh S.|last2 = Kumar|first3 = Alex Pappachen|last3 = James|doi-access = free}}</ref> and [[object recognition]].<ref>{{Cite journal|title = Memristive Threshold Logic Circuit Design of Fast Moving Object Detection|journal = IEEE Transactions on Very Large Scale Integration (VLSI) Systems|date = 2015-10-01|issn = 1063-8210|pages = 2337–2341|volume = 23|issue = 10|doi = 10.1109/TVLSI.2014.2359801|first1 = A.K.|last1 = Maan|first2 = D.S.|last2 = Kumar|first3 = S.|last3 = Sugathan|first4 = A.P.|last4 = James|arxiv = 1410.1267|s2cid = 9647290}}</ref> They also find applications in replacing conventional digital logic gates.<ref>{{Cite journal|title = Resistive Threshold Logic|journal = IEEE Transactions on Very Large Scale Integration (VLSI) Systems|date = 2014-01-01|issn = 1063-8210|pages = 190–195|volume = 22|issue = 1|doi = 10.1109/TVLSI.2012.2232946|first1 = A.P.|last1 = James|first2 = L.R.V.J.|last2 = Francis|first3 = D.S.|last3 = Kumar|arxiv = 1308.0090|s2cid = 7357110}}</ref><ref>{{Cite journal|title = Threshold Logic Computing: Memristive-CMOS Circuits for Fast Fourier Transform and Vedic Multiplication|journal = IEEE Transactions on Very Large Scale Integration (VLSI) Systems|date = 2015-11-01|issn = 1063-8210|pages = 2690–2694|volume = 23|issue = 11|doi = 10.1109/TVLSI.2014.2371857|first1 = A.P.|last1 = James|first2 = D.S.|last2 = Kumar|first3 = A.|last3 = Ajayan|arxiv = 1411.5255|s2cid = 6076956}}</ref><br />
<br />
There exist several neuron inspired threshold logic functions implemented with memristors that have applications in high level pattern recognition applications. Some of the applications reported recently include speech recognition, face recognition and object recognition. They also find applications in replacing conventional digital logic gates.<br />
<br />
利用记忆电阻器实现的神经元启发阈值逻辑函数在高级模式识别中有着广泛的应用。最近报道的一些应用包括语音识别、人脸识别和物体识别。它们还可以用来取代传统的数字逻辑门。<br />
<br />
For ideal passive memristive circuits there is an exact equation (Caravelli-Traversa-[[Di Ventra]] equation) for the internal memory of the circuit:<ref>{{cite journal |last=Caravelli |display-authors=etal|arxiv=1608.08651 |title=The complex dynamics of memristive circuits: analytical results and universal slow relaxation |year=2017 |doi=10.1103/PhysRevE.95.022140 |pmid= 28297937 |volume=95 |issue= 2 |pages= 022140 |journal=Physical Review E|bibcode=2017PhRvE..95b2140C |s2cid=6758362}}</ref><br />
<br />
For ideal passive memristive circuits there is an exact equation (Caravelli-Traversa-Di Ventra equation) for the internal memory of the circuit:<br />
<br />
对于理想的无源记忆电路，电路的内部记忆有一个精确的方程(Caravelli-Traversa-Di Ventra 方程) :<br />
<br />
:<math> \frac{d}{dt} \vec{W} = \alpha \vec{W}-\frac{1}{\beta} (I+\xi \Omega W)^{-1} \Omega \vec S </math><br />
<br />
:\frac{d}{dt} \vec{W} = \alpha \vec{W}-\frac{1}{\beta} (I+\xi \Omega W)^{-1} \Omega \vec S<br />
<br />
:\frac{d}{dt} \vec{W} = \alpha \vec{W}-\frac{1}{\beta} (I+\xi \Omega W)^{-1} \Omega \vec S<br />
<br />
as a function of the properties of the physical memristive network and the external sources. In the equation above, <math>\alpha</math> is the "forgetting" time scale constant, <math> \xi=r-1</math> and <math>r=\frac{R_\text{off}}{R_\text{on}}</math> is the ratio of ''off'' and ''on'' values of the limit resistances of the memristors, <math> \vec S </math> is the vector of the sources of the circuit and <math>\Omega</math> is a projector on the fundamental loops of the circuit. The constant <math>\beta</math> has the dimension of a voltage and is associated to the properties of the [[memristor]]; its physical origin is the charge mobility in the conductor. The diagonal matrix and vector <math>W=\operatorname{diag}(\vec W)</math> and <math>\vec W</math> respectively, are instead the internal value of the memristors, with values between 0 and 1. This equation thus requires adding extra constraints on the memory values in order to be reliable.<br />
<br />
<nowiki>as a function of the properties of the physical memristive network and the external sources. In the equation above, \alpha is the "forgetting" time scale constant, \xi=r-1 and r=\frac{R_\text{off}}{R_\text{on}} is the ratio of off and on values of the limit resistances of the memristors, \vec S is the vector of the sources of the circuit and \Omega is a projector on the fundamental loops of the circuit. The constant \beta has the dimension of a voltage and is associated to the properties of the memristor; its physical origin is the charge mobility in the conductor. The diagonal matrix and vector W=\operatorname{diag}(\vec W) and \vec W respectively, are instead the internal value of the memristors, with values between 0 and 1. This equation thus requires adding extra constraints on the memory values in order to be reliable.</nowiki><br />
<br />
<nowiki>作为物理记忆网络和外部源的性质的函数。在上述方程中，α 是“遗忘”时间尺度常数，xi = r-1，r = frac { r _ text { off }{ on }{ r _ text { on }}是记忆电阻器极限电阻的开关和开关值之比，vec s 是电路源的矢量，Omega 是电路基本环路的投影仪。常数 β 具有电压的尺寸，与记忆电阻器的特性有关; 它的物理起源是导体中的电荷迁移率。对角矩阵和向量 w = 操作者名{ diag }(vec w)和 vec w 分别是记忆电阻器的内值，值在0到1之间。因此，这个等式需要在内存值上添加额外的约束，以保证可靠性。</nowiki><br />
<br />
==See also ==<br />
==相关词条==<br />
{{Columns-list|colwidth=18em|<br />
* [[AI accelerator (computer hardware)|AI accelerator]]<br />
* [[Artificial brain]]<br />
* [[Biomorphic]]<br />
* [[Cognitive computer]]<br />
* [[Computation and Neural Systems]]<br />
* [[Differentiable programming]]<br />
* [[Event camera]]<br />
* [[Neurorobotics]]<br />
* [[Optical flow sensor]]<br />
* [[Physical neural network]]<br />
* [[SpiNNaker]]<br />
* [[SyNAPSE]]<br />
* [[Retinomorphic sensor]]<br />
* [[Vision chip]]<br />
* [[Vision processing unit]]<br />
* [[Zeroth (software)]]<br />
* [[Hardware for artificial intelligence]]<br />
}}<br />
<br />
{{Portal bar|Electronics}}<br />
<br />
==References==<br />
{{Reflist|40em}}<br />
<br />
==External links==<br />
<!--======================== {{No more links}} ============================<br />
| PLEASE BE CAUTIOUS IN ADDING MORE LINKS TO THIS ARTICLE. Wikipedia |<br />
| is not a collection of links nor should it be used for advertising. |<br />
| |<br />
| Excessive or inappropriate links WILL BE DELETED. |<br />
| See [[Wikipedia:External links]] & [[Wikipedia:Spam]] for details. |<br />
| |<br />
| If there are already plentiful links, please propose additions or |<br />
| replacements on this article's discussion page, or submit your link |<br />
| to the relevant category at the Open Directory Project (dmoz.org) |<br />
| and link back to that category using the {{dmoz}} template. |<br />
======================={{No more links}}=============================--><br />
{{Commons category|Neuromorphic Engineering}}<br />
*[https://web.archive.org/web/20150727034331/http://ine-web.org/workshops/workshops-overview Telluride Neuromorphic Engineering Workshop]<br />
*[https://archive.today/20130115190057/http://capocaccia.ethz.ch/ CapoCaccia Cognitive Neuromorphic Engineering Workshop]<br />
*[https://web.archive.org/web/20190716132350/http://www.ine-web.org/ Institute of Neuromorphic Engineering]<br />
*[http://www.ine-news.org/ INE news site].<br />
*[http://www.frontiersin.org/neuromorphic_engineering Frontiers in Neuromorphic Engineering Journal]<br />
*[http://www.cns.caltech.edu/ Computation and Neural Systems] department at the [[California Institute of Technology]].<br />
*[http://www.humanbrainproject.eu/ Human Brain Project official site]<br />
*[https://www.the-scientist.com/features/building-a-silicon-brain-65738 Building a Silicon Brain:] Computer chips based on biological neurons may help simulate larger and more-complex brain models. May 1, 2019. SANDEEP RAVINDRAN<br />
<br />
<br />
*Telluride Neuromorphic Engineering Workshop<br />
*CapoCaccia Cognitive Neuromorphic Engineering Workshop<br />
*Institute of Neuromorphic Engineering<br />
*INE news site.<br />
*Frontiers in Neuromorphic Engineering Journal<br />
*Computation and Neural Systems department at the California Institute of Technology.<br />
*Human Brain Project official site<br />
*Building a Silicon Brain: Computer chips based on biological neurons may help simulate larger and more-complex brain models. May 1, 2019. SANDEEP RAVINDRAN<br />
<br />
=<nowiki>外部链接</nowiki>= <br />
* 碲化物神经形态工程工作室<br />
*CapoCaccia 认知神经形态工程工作室<br />
* 神经形态工程研究所<br />
*INE 新闻站点。<br />
*《神经形态工程学前沿》<br />
*加州理工学院计算与神经系统系。<br />
*人脑项目官方网站<br />
*打造硅脑: 基于生物神经元的计算机芯片可能有助于模拟更大、更复杂的大脑模型。2019年5月1日。SANDEEP RAVINDRAN<br />
<br />
{{Differentiable computing}}<br />
{{Authority control}}<br />
<br />
[[Category:Electrical engineering]]<br />
[[Category:Neuroscience]]<br />
[[分类:AI accelerators|*]]<br />
[[Category:Artificial intelligence]]<br />
[[Category:Robotics]]<br />
<br />
Category:Electrical engineering<br />
Category:Neuroscience<br />
*<br />
Category:Artificial intelligence<br />
Category:Robotics<br />
<br />
类别: 电气工程类别: 神经科学 * 类别: 人工智能类别: 机器人<br />
<br />
<noinclude><br />
<br />
<small>This page was moved from [[wikipedia:en:Neuromorphic engineering]]. Its edit history can be viewed at [[神经形态计算/edithistory]]</small></noinclude><br />
<br />
[[Category:待整理页面]]<br />
<br />
<references /></div>Shenky20https://wiki.swarma.org/index.php?title=%E7%A5%9E%E7%BB%8F%E5%BD%A2%E6%80%81%E8%AE%A1%E7%AE%97&diff=30177神经形态计算2022-04-13T08:25:00Z<p>Shenky20：/* 神经生物学启发 */</p>
<hr />
<div><br />
此词条由神经动力学模型读书会词条梳理志愿者Shenky20翻译审校，未经专家审校，带来阅读不便，请见谅。<br />
<br />
{{Use American English|date = January 2019}}<br />
{{Use mdy dates|date = January 2019}}<br />
<br />
'''Neuromorphic engineering''', also known as '''neuromorphic computing''',<ref name=":3">{{Cite journal | doi = 10.1145/2601069| title = Neuromorphic computing gets ready for the (really) big time| journal = [[Communications of the ACM]]| volume = 57| issue = 6| year = 2014| last1 = Monroe | first1 = D. | pages = 13–15| s2cid = 20051102}}</ref><ref name=":4">{{Cite journal | doi = 10.1088/0957-4484/21/17/175202| pmid = 20368686| title = Nanotube devices based crossbar architecture: Toward neuromorphic computing| journal = Nanotechnology| volume = 21| issue = 17| pages = 175202| year = 2010| last1 = Zhao | first1 = W. S. | last2 = Agnus | first2 = G. | last3 = Derycke | first3 = V. | last4 = Filoramo | first4 = A. | last5 = Bourgoin | first5 = J. -P. | last6 = Gamrat | first6 = C. | bibcode = 2010Nanot..21q5202Z| url = https://zenodo.org/record/3428659}}</ref><ref name="humanbrainproject">{{YouTube|id=6RoiZ90mGfw|title=The Human Brain Project SP 9: Neuromorphic Computing Platform}}</ref> is the use of [[very-large-scale integration]] (VLSI) systems containing electronic [[analog circuit]]s to mimic neuro-biological architectures present in the nervous system. A neuromorphic computer/chip is any device that uses physical artificial neurons (made from silicon) to do computations.<ref name=":5">{{cite journal|last1=Mead|first1=Carver|title=Neuromorphic electronic systems|journal=Proceedings of the IEEE|date=1990|volume=78|issue=10|pages=1629–1636|doi=10.1109/5.58356|url=https://authors.library.caltech.edu/53090/1/00058356.pdf}}</ref><ref name=":2" /> In recent times, the term ''neuromorphic'' has been used to describe analog, digital, mixed-mode analog/digital VLSI, and software systems that implement models of [[neural system]]s (for [[perception]], [[motor control]], or [[multisensory integration]]). The implementation of neuromorphic computing on the hardware level can be realized by oxide-based [[memristor]]s,<ref name="Maan 1–13">{{Cite journal|last1=Maan|first1=A. K.|last2=Jayadevi|first2=D. A.|last3=James|first3=A. P.|date=2016-01-01|title=A Survey of Memristive Threshold Logic Circuits|journal=IEEE Transactions on Neural Networks and Learning Systems|volume=PP|issue=99|pages=1734–1746|doi=10.1109/TNNLS.2016.2547842|pmid=27164608|issn=2162-237X|arxiv=1604.07121|bibcode=2016arXiv160407121M|s2cid=1798273}}</ref> spintronic memories, threshold switches, and [[transistor]]s.<ref name=":6">{{Cite journal|title = Mott Memory and Neuromorphic Devices|journal = Proceedings of the IEEE|date = 2015-08-01|issn = 0018-9219|pages = 1289–1310|volume = 103|issue = 8|doi = 10.1109/JPROC.2015.2431914|first1 = You|last1 = Zhou|first2 = S.|last2 = Ramanathan|s2cid = 11347598|url = https://zenodo.org/record/895565}}</ref><ref name=":2">{{Cite document|title=Neuromorphic Circuits With Neural Modulation Enhancing the Information Content of Neural Signaling {{!}} International Conference on Neuromorphic Systems 2020|language=EN|doi=10.1145/3407197.3407204|s2cid=220794387}}</ref> Training software-based neuromorphic systems of spiking neural networks can be achieved using error backpropagation, e.g., using [[Python (programming language)|Python]] based frameworks such as snnTorch,<ref name=":7">{{cite journal |last1=Eshraghian|first1=Jason K.|last2=Ward|first2=Max|last3=Neftci |first3=Emre|last4=Wang|first4=Xinxin|last5=Lenz|first5=Gregor|last6=Dwivedi|first6=Girish|last7=Bennamoun|first7=Mohammed|last8=Jeong|first8=Doo Seok|last9=Lu|first9=Wei D.|title=Training Spiking Neural Networks Using Lessons from Deep Learning |date=1 October 2021 |arxiv=2109.12894 }}</ref> or using canonical learning rules from the biological learning literature, e.g., using BindsNet.<ref name=":8">{{Cite web | url=https://github.com/Hananel-Hazan/bindsnet | title=Hananel-Hazan/bindsnet: Simulation of spiking neural networks (SNNs) using PyTorch.| date=31 March 2020}}</ref><br />
<br />
<br />
'''<font color="#ff8000">神经形态工程Neuromorphic engineering</font>'''(也称为'''<font color="#ff8000">神经形态计算Neuromorphic computing</font>''')<ref name=":3" /><ref name=":4" /><ref name="humanbrainproject" />是指使用包含电子'''<font color="#ff8000">模拟电路Analog circuit</font>'''的'''<font color="#ff8000">超大规模集成电路Very-large-scale integration</font>'''系统来模拟神经系统中生理结构的研究方法。神经形态计算机或神经形态芯片包括任何使用由硅制成的人造神经元进行计算的设备。<ref name=":5" /><ref name=":2" />近年来，神经形态学（neuromorphic）这个术语被用来描述能够实现'''<font color="#ff8000">神经系统Neural system</font>'''模型功能(如'''<font color="#ff8000">感知Perception</font>'''、'''<font color="#ff8000">运动控制Motor control</font>'''，'''<font color="#ff8000">多感官整合Multisensory integration</font>'''等)的模拟、数字、模拟/数字混合模式超大规模集成电路和软件系统。神经形态计算的硬件实现可以通过基于氧化物的'''<font color="#ff8000">记忆电阻器Memristor</font>'''、<ref name="Maan 1–13" />自旋电子存储器、阈值开关和'''<font color="#ff8000">晶体管Transistor</font>'''来实现。<ref name=":6" /><ref name=":2" />对基于软件的脉冲神经网络系统的训练可以通过误差反向传播机制来实现，例如，使用snnTorch等基于Python的框架，<ref name=":7" />或使用BindsNet等典型的受生物启发的学习模式。<ref name=":8" /><br />
<br />
A key aspect of neuromorphic engineering is understanding how the morphology of individual neurons, circuits, applications, and overall architectures creates desirable computations, affects how information is represented, influences robustness to damage, incorporates learning and development, adapts to local change (plasticity), and facilitates evolutionary change.<br />
<br />
<br />
神经形态工程领域的一个关键问题，就是理解'''<font color="#32CD32">单个神经元形态、神经回路、应用和整体结构</font>'''如何产生理想的计算，如何影响信息的表示和对破坏的鲁棒性，如何整合学习和发展，如何适应局部变化(可塑性) 并促进逐渐发展的变化。<br />
<br />
Neuromorphic engineering is an interdisciplinary subject that takes inspiration from [[biology]], [[physics]], [[mathematics]], [[computer science]], and [[electronic engineering]]<ref name=":2" /> to design artificial neural systems, such as [[Machine vision|vision systems]], head-eye systems, auditory processors, and autonomous robots, whose physical architecture and design principles are based on those of biological nervous systems.<ref name=":9">{{Cite journal | doi = 10.1155/2012/705483| title = Qualitative Functional Decomposition Analysis of Evolved Neuromorphic Flight Controllers| journal = Applied Computational Intelligence and Soft Computing| volume = 2012| pages = 1–21| year = 2012| last1 = Boddhu | first1 = S. K. | last2 = Gallagher | first2 = J. C. | doi-access = free}}</ref> It was developed by [[Carver Mead]]<ref name=":10">{{cite web|title=carver mead website|last1=Mead|first1=Carver|url=http://www.carvermead.caltech.edu/index.html|website=carvermead}}</ref> in the late 1980s.<br />
<br />
<br />
神经形态工程是以'''<font color="#ff8000">生物学Biology</font>'''、'''<font color="#ff8000">物理学Physics</font>'''、'''<font color="#ff8000">数学Mathematics</font>'''、'''<font color="#ff8000">计算机科学Computer science</font>'''和'''<font color="#ff8000">电子工程Electronic engineering<ref name=":2" /></font>'''等学科为基础，设计人工神经系统（如'''<font color="#ff8000">视觉系统Vision systems</font>'''、头眼系统、听觉处理器以及物理结构和设计原则都受启发于生物神经系统的自主机器人）的一门交叉学科。<ref name=":9" />20世纪80年代后期，卡弗·米德极大地推动了神经形态工程领域的发展。<ref name=":10" /><br />
<br />
==Neurological inspiration==<br />
==神经生物学启发==<br />
Neuromorphic engineering is set apart by the inspiration it takes from what we know about the structure and operations of the [[brain]]. Neuromorphic engineering translates what we know about the brain's function into computer systems. Work has mostly focused on replicating the analog nature of [[biological computation]] and the role of [[neuron]]s in [[cognition]].<br />
<br />
<br />
神经形态工程形成的灵感来源于现已知的'''<font color="#ff8000">大脑Brain</font>'''结构及运作机制知识，它将我们对大脑功能的了解用于研究和优化计算机系统。该领域工作主要集中于对'''<font color="#ff8000">生物计算Biological computation</font>'''模拟特性和'''<font color="#ff8000">神经元Neuron</font>'''在'''<font color="#ff8000">认知Cognition</font>'''中所发挥作用的复刻。<br />
<br />
The biological processes of neurons and their [[synapse]]s are dauntingly complex, and thus very difficult to artificially simulate. A key feature of biological brains is that all of the processing in neurons use analog chemical signals. This makes it hard to replicate brains in computers because the current generation of computers is completely digital. However, the characteristics of these parts can be abstracted into mathematical functions that closely capture the essence of the neuron's operations.<br />
<br />
神经元及其'''<font color="#ff8000">突触Synapse</font>'''的生理学过程极其复杂，难以进行人工模拟。神经元在所有处理过程中都使用化学模拟信号，这是大脑的一个关键生理学特征。这个特征大大增加了在计算机中复制大脑的难度，因为目前的计算机是完全数字化的。然而，部分特征可以抽象为数学函数，这些函数能够紧密捕捉神经元运行的本质。<br />
<br />
The goal of neuromorphic computing is not to perfectly mimic the brain and all of its functions, but instead to extract what is known of its structure and operations to be used in a practical computing system. No neuromorphic system will claim nor attempt to reproduce every element of neurons and synapses, but all adhere to the idea that computation is highly [[distributed processing|distributed]] throughout a series of small computing elements analogous to a neuron. While this sentiment is standard, researchers chase this goal with different methods.<ref>{{Cite journal | doi = 10.1088/1741-2560/13/5/051001| title = Large-scale neuromorphic computing systems| journal = Journal of Neural Engineering| volume = 13| pages = 1–15| year = 2016| last1 = Furber | first1 = Steve| issue = 5| pmid = 27529195| bibcode = 2016JNEng..13e1001F| doi-access = free}}</ref><br />
<br />
The goal of neuromorphic computing is not to perfectly mimic the brain and all of its functions, but instead to extract what is known of its structure and operations to be used in a practical computing system. No neuromorphic system will claim nor attempt to reproduce every element of neurons and synapses, but all adhere to the idea that computation is highly distributed throughout a series of small computing elements analogous to a neuron. While this sentiment is standard, researchers chase this goal with different methods.<br />
<br />
神经形态计算的目标不是完美地模拟大脑及其所有功能，而是利用已知的大脑结构和运转机制来研发或优化实际的计算系统。任何神经形态学系统都不会声称或试图复制神经元和突触中的每一个元素，但所有人都一致认可将计算高度分散于一系列类似于神经元的小型计算元素的理念。研究人员用不同的方法来追求这一普遍目标。<br />
<br />
<br />
==Examples==<br />
==案例 ==<br />
As early as 2006, researchers at [[Georgia Tech]] published a field programmable neural array.<ref>{{Cite book|title = A field programmable neural array|last1 = Farquhar|first1 = Ethan|date = May 2006|journal = IEEE International Symposium on Circuits and Systems|pages = 4114–4117|last2 = Hasler|first2 = Paul.|doi = 10.1109/ISCAS.2006.1693534|isbn = 978-0-7803-9389-9|s2cid = 206966013}}</ref> This chip was the first in a line of increasingly complex arrays of floating gate transistors that allowed programmability of charge on the gates of [[MOSFET]]s to model the channel-ion characteristics of neurons in the brain and was one of the first cases of a silicon programmable array of neurons.<br />
<br />
As early as 2006, researchers at Georgia Tech published a field programmable neural array. This chip was the first in a line of increasingly complex arrays of floating gate transistors that allowed programmability of charge on the gates of MOSFETs to model the channel-ion characteristics of neurons in the brain and was one of the first cases of a silicon programmable array of neurons.<br />
<br />
早在2006年，佐治亚理工学院的研究人员就发表了现场可编程神经阵列。这种芯片是一系列越来越复杂的浮栅晶体管阵列中的第一个，这些晶体管可以在 mosfet 的栅极上编程来模拟大脑中神经元的通道离子特性，这也是硅可编程神经元阵列的首批案例之一。<br />
<br />
In November 2011, a group of [[MIT]] researchers created a computer chip that mimics the analog, ion-based communication in a synapse between two neurons using 400 transistors and standard [[CMOS]] manufacturing techniques.<ref>{{cite web|title=MIT creates "brain chip"|url=http://www.extremetech.com/extreme/105067-mit-creates-brain-chip|access-date=4 December 2012}}</ref><ref name="Neuromorphic silicon paper">{{cite journal|title=Neuromorphic silicon neurons and large-scale neural networks: challenges and opportunities|doi=10.3389/fnins.2011.00108|pmid=21991244|pmc=3181466|volume=5|pages=108|journal=Frontiers in Neuroscience|year=2011|last1=Poon|first1=Chi-Sang|last2=Zhou|first2=Kuan|doi-access=free}}</ref><br />
<br />
In November 2011, a group of MIT researchers created a computer chip that mimics the analog, ion-based communication in a synapse between two neurons using 400 transistors and standard CMOS manufacturing techniques.<br />
<br />
2011年11月，麻省理工学院的一组研究人员发明了一种计算机芯片，该芯片使用400个晶体管和标准的 CMOS 制造技术，在两个神经元之间的突触中模拟模拟基于离子的通讯。<br />
<br />
In June 2012, [[spintronic]] researchers at [[Purdue University]] presented a paper on the design of a neuromorphic chip using [[Spin valve|lateral spin valve]]s and [[memristor]]s. They argue that the architecture works similarly to neurons and can therefore be used to test methods of reproducing the brain's processing. In addition, these chips are significantly more energy-efficient than conventional ones.<ref name="Spin Devices Prop">{{Cite arXiv|title=Proposal For Neuromorphic Hardware Using Spin Devices|eprint=1206.3227|last1=Sharad|first1=Mrigank|last2=Augustine|first2=Charles|last3=Panagopoulos|first3=Georgios|last4=Roy|first4=Kaushik|class=cond-mat.dis-nn|year=2012}}</ref><br />
<br />
In June 2012, spintronic researchers at Purdue University presented a paper on the design of a neuromorphic chip using lateral spin valves and memristors. They argue that the architecture works similarly to neurons and can therefore be used to test methods of reproducing the brain's processing. In addition, these chips are significantly more energy-efficient than conventional ones.<br />
<br />
2012年6月，普渡大学的自旋电子学研究人员发表了一篇关于利用侧向自旋阀和记忆电阻器设计神经形态芯片的论文。他们认为，这种结构的工作原理与神经元相似，因此可以用来测试复制大脑处理过程的方法。此外，这些芯片明显比传统芯片更节能。<br />
<br />
Research at [[HP Labs]] on Mott memristors has shown that while they can be non-[[Volatile memory|volatile]], the volatile behavior exhibited at temperatures significantly below the [[phase transition]] temperature can be exploited to fabricate a [[neuristor]],<ref name=":0" /> a biologically-inspired device that mimics behavior found in neurons.<ref name=":0">{{Cite journal | doi = 10.1038/nmat3510| pmid = 23241533| title = A scalable neuristor built with Mott memristors| journal = Nature Materials| volume = 12| issue = 2| pages = 114–7| year = 2012| last1 = Pickett | first1 = M. D. | last2 = Medeiros-Ribeiro | first2 = G. | last3 = Williams | first3 = R. S. | bibcode = 2013NatMa..12..114P| s2cid = 16271627| url = https://semanticscholar.org/paper/b6ba6f496ace2b947f111059663e76bb60e9efeb}}</ref> In September 2013, they presented models and simulations that show how the spiking behavior of these neuristors can be used to form the components required for a [[Turing machine]].<ref>{{cite journal|doi=10.1088/0957-4484/24/38/384002|title=Phase transitions enable computational universality in neuristor-based cellular automata|author1=Matthew D Pickett|author2=R Stanley Williams|name-list-style=amp|date=September 2013|publisher=IOP Publishing Ltd|journal=Nanotechnology|volume=24|issue=38|pmid=23999059|bibcode=2013Nanot..24L4002P|at=384002}}</ref><br />
<br />
Research at HP Labs on Mott memristors has shown that while they can be non-volatile, the volatile behavior exhibited at temperatures significantly below the phase transition temperature can be exploited to fabricate a neuristor, a biologically-inspired device that mimics behavior found in neurons. In September 2013, they presented models and simulations that show how the spiking behavior of these neuristors can be used to form the components required for a Turing machine.<br />
<br />
惠普实验室在莫特记忆电阻器上的研究表明，尽管它们可以是非挥发性的，但是在相变温度以下显示的挥发性行为可以被用来制造神经元电阻器，这是一种模仿神经元行为的生物设备。2013年9月，他们展示了模型和仿真，展示了这些神经元的尖峰行为是如何被用来形成图灵机所需的元件的。<br />
<br />
[[Neurogrid]], built by ''Brains in Silicon'' at [[Stanford University]],<ref>{{cite journal|last1=Boahen|first1=Kwabena|title=Neurogrid: A Mixed-Analog-Digital Multichip System for Large-Scale Neural Simulations|journal=Proceedings of the IEEE|date=24 April 2014|volume=102|issue=5|pages=699–716|doi=10.1109/JPROC.2014.2313565|s2cid=17176371}}</ref> is an example of hardware designed using neuromorphic engineering principles. The circuit board is composed of 16 custom-designed chips, referred to as NeuroCores. Each NeuroCore's analog circuitry is designed to emulate neural elements for 65536 neurons, maximizing energy efficiency. The emulated neurons are connected using digital circuitry designed to maximize spiking throughput.<ref>{{cite journal|doi=10.1038/503022a|pmid = 24201264|title = Neuroelectronics: Smart connections|journal = Nature|volume = 503|issue = 7474|pages = 22–4|year = 2013|last1 = Waldrop|first1 = M. Mitchell|bibcode = 2013Natur.503...22W|doi-access = free}}</ref><ref>{{cite journal|doi=10.1109/JPROC.2014.2313565|title = Neurogrid: A Mixed-Analog-Digital Multichip System for Large-Scale Neural Simulations|journal = Proceedings of the IEEE|volume = 102|issue = 5|pages = 699–716|year = 2014|last1 = Benjamin|first1 = Ben Varkey|last2 = Peiran Gao|last3 = McQuinn|first3 = Emmett|last4 = Choudhary|first4 = Swadesh|last5 = Chandrasekaran|first5 = Anand R.|last6 = Bussat|first6 = Jean-Marie|last7 = Alvarez-Icaza|first7 = Rodrigo|last8 = Arthur|first8 = John V.|last9 = Merolla|first9 = Paul A.|last10 = Boahen|first10 = Kwabena|s2cid = 17176371}}</ref><br />
<br />
Neurogrid, built by Brains in Silicon at Stanford University, is an example of hardware designed using neuromorphic engineering principles. The circuit board is composed of 16 custom-designed chips, referred to as NeuroCores. Each NeuroCore's analog circuitry is designed to emulate neural elements for 65536 neurons, maximizing energy efficiency. The emulated neurons are connected using digital circuitry designed to maximize spiking throughput.<br />
<br />
神经网格是由斯坦福大学硅谷大脑公司建立的，是一个使用神经形态工程原理设计硬件的例子。该电路板由16个定制设计的芯片组成，称为 NeuroCores。每个 NeuroCore 的模拟电路被设计为模拟65536个神经元的神经元元件，最大限度地提高能量效率。模拟的神经元通过设计的数字电路连接，以最大化脉冲吞吐量。<br />
<br />
A research project with implications for neuromorphic engineering is the [[Human Brain Project]] that is attempting to simulate a complete human brain in a supercomputer using biological data. It is made up of a group of researchers in neuroscience, medicine, and computing.<ref>{{cite web|title=Involved Organizations|url=http://www.humanbrainproject.eu/partners.html|access-date=22 February 2013|url-status=dead|archive-url=https://web.archive.org/web/20130302142627/http://www.humanbrainproject.eu/partners.html|archive-date=2 March 2013}}</ref> [[Henry Markram]], the project's co-director, has stated that the project proposes to establish a foundation to explore and understand the brain and its diseases, and to use that knowledge to build new computing technologies. The three primary goals of the project are to better understand how the pieces of the brain fit and work together, to understand how to objectively diagnose and treat brain diseases, and to use the understanding of the human brain to develop neuromorphic computers. That the simulation of a complete human brain will require a supercomputer a thousand times more powerful than today's encourages the current focus on neuromorphic computers.<ref>{{cite web|title=Human Brain Project|url=http://www.humanbrainproject.eu|access-date=22 February 2013}}</ref> $1.3 billion has been allocated to the project by The [[European Commission]].<ref>{{cite web|title=The Human Brain Project and Recruiting More Cyberwarriors|url=http://www.marketplace.org/topics/tech/human-brain-project-and-recruiting-more-cyberwarriors|access-date=22 February 2013|date=January 29, 2013}}</ref><br />
<br />
A research project with implications for neuromorphic engineering is the Human Brain Project that is attempting to simulate a complete human brain in a supercomputer using biological data. It is made up of a group of researchers in neuroscience, medicine, and computing. Henry Markram, the project's co-director, has stated that the project proposes to establish a foundation to explore and understand the brain and its diseases, and to use that knowledge to build new computing technologies. The three primary goals of the project are to better understand how the pieces of the brain fit and work together, to understand how to objectively diagnose and treat brain diseases, and to use the understanding of the human brain to develop neuromorphic computers. That the simulation of a complete human brain will require a supercomputer a thousand times more powerful than today's encourages the current focus on neuromorphic computers. $1.3 billion has been allocated to the project by The European Commission.<br />
<br />
一个对神经形态工程有影响的研究项目是人脑项目，它试图用生物数据在超级计算机中模拟完整的人脑。它由一群神经科学、医学和计算机科学的研究人员组成。该项目的联合主管亨利•马克拉姆(Henry Markram)表示，该项目提议建立一个基础，以探索和了解大脑及其疾病，并利用这些知识构建新的计算机技术。这个项目的三个主要目标是: 更好地理解大脑的各个部分是如何相互匹配和协同工作的; 理解如何客观地诊断和治疗脑部疾病; 以及利用对人类大脑的理解来开发神经形态计算机。模拟一个完整的人类大脑需要一台比现在强大一千倍的超级计算机，这鼓励了当前对神经形态计算机的关注。欧洲委员会已经拨款13亿美元用于这个项目。<br />
<br />
Other research with implications for neuromorphic engineering involves the [[BRAIN Initiative]]<ref name="economist">[https://www.economist.com/news/science-and-technology/21582495-computers-will-help-people-understand-brains-better-and-understanding-brains Neuromorphic computing: The machine of a new soul], The Economist, 2013-08-03</ref> and the [[TrueNorth]] chip from [[IBM]].<ref>{{cite journal|last1=Modha|first1=Dharmendra|title=A million spiking-neuron integrated circuit with a scalable communication network and interface|journal=Science|date=Aug 2014|volume=345|issue=6197|pages=668–673|doi=10.1126/science.1254642|pmid=25104385|bibcode=2014Sci...345..668M|s2cid=12706847}}</ref> Neuromorphic devices have also been demonstrated using nanocrystals, nanowires, and conducting polymers.<ref>{{Cite web|url=http://jessamynfairfield.com/wp-content/uploads/2017/03/PWMar17Fairfield.pdf|title=Smarter Machines|last=Fairfield|first=Jessamyn|date=March 1, 2017}}</ref><br />
<br />
Other research with implications for neuromorphic engineering involves the BRAIN InitiativeNeuromorphic computing: The machine of a new soul, The Economist, 2013-08-03 and the TrueNorth chip from IBM. Neuromorphic devices have also been demonstrated using nanocrystals, nanowires, and conducting polymers.<br />
<br />
其他与神经形态工程有关的研究还包括 BRAIN initiativity/euromorphic computing: The machine of a new soul，The Economist，2013-08-03 and The TrueNorth chip from IBM。神经形态学设备也已经被证明使用纳米晶体、纳米线和导电聚合物。<br />
<br />
[[Intel]] unveiled its neuromorphic research chip, called “[[Intel Loihi|Loihi]]”, in October 2017. The chip uses an asynchronous [[spiking neural network]] (SNN) to implement adaptive self-modifying event-driven fine-grained parallel computations used to implement learning and inference with high efficiency.<ref>{{cite journal |last1=Davies |first1=Mike |title=Loihi: A Neuromorphic Manycore Processor with On-Chip Learning |journal=IEEE Micro |date=January 16, 2018 |volume=38 |issue=1 |pages=82–99 |display-authors=etal|doi=10.1109/MM.2018.112130359 |s2cid=3608458 }}</ref><ref name="Morris2017">{{cite web |last1=Morris |first1=John |title=Why Intel built a neuromorphic chip |url=https://www.zdnet.com/article/why-intel-built-a-neuromorphic-chip/ |website=ZDNet |access-date=17 August 2018 |language=en}}</ref><br />
<br />
Intel unveiled its neuromorphic research chip, called “Loihi”, in October 2017. The chip uses an asynchronous spiking neural network (SNN) to implement adaptive self-modifying event-driven fine-grained parallel computations used to implement learning and inference with high efficiency.<br />
<br />
2017年10月，英特尔公布了其神经形态研究芯片“ Loihi”。该芯片采用异步脉冲神经网络(SNN)实现自适应自修改事件驱动的细粒度并行计算，实现了高效的学习和推理。<br />
<br />
[[IMEC]], a Belgium-based nanoelectronics research center, demonstrated the world's first self-learning neuromorphic chip. The brain-inspired chip, based on OxRAM technology, has the capability of self-learning and has been demonstrated to have the ability to compose music.<ref>{{cite web |title=Imec demonstrates self-learning neuromorphic chip that composes music |url=https://www.imec-int.com/en/articles/imec-demonstrates-self-learning-neuromorphic-chip-that-composes-music |website=IMEC International |access-date=1 October 2019}}</ref> IMEC released the 30-second tune composed by the prototype. The chip was sequentially loaded with songs in the same time signature and style. The songs were old Belgian and French flute minuets, from which the chip learned the rules at play and then applied them.<ref>{{cite web|last1=Bourzac|first1=Katherine|title=A Neuromorphic Chip That Makes Music|url=https://spectrum.ieee.org/a-neuromorphic-chip-that-makes-music|url-status=live|access-date=1 October 2019|website=IEEE Spectrum|date=May 23, 2017}}</ref><br />
<br />
IMEC, a Belgium-based nanoelectronics research center, demonstrated the world's first self-learning neuromorphic chip. The brain-inspired chip, based on OxRAM technology, has the capability of self-learning and has been demonstrated to have the ability to compose music. IMEC released the 30-second tune composed by the prototype. The chip was sequentially loaded with songs in the same time signature and style. The songs were old Belgian and French flute minuets, from which the chip learned the rules at play and then applied them.<br />
<br />
比利时的一个纳米电子学研究中心 IMEC 展示了世界上第一个自学习神经形态芯片。这种基于 OxRAM 技术的大脑启发芯片具有自学习能力，并已被证明具有创作音乐的能力。IMEC 发布了由原型机谱写的30秒曲调。芯片按顺序加载同时签名和风格的歌曲。歌曲是古老的比利时和法国长笛小步舞曲，筹码从中学习游戏规则，然后应用它们。<br />
<br />
[[Blue Brain Project|The Blue Brain Project]], led by Henry Markram, aims to build biologically detailed digital reconstructions and simulations of the mouse brain. The Blue Brain Project has created in silico models of rodent brains, while attempting to replicate as many details about its biology as possible. The supercomputer-based simulations offer new perspectives on understanding the structure and functions of the brain.<br />
<br />
The Blue Brain Project, led by Henry Markram, aims to build biologically detailed digital reconstructions and simulations of the mouse brain. The Blue Brain Project has created in silico models of rodent brains, while attempting to replicate as many details about its biology as possible. The supercomputer-based simulations offer new perspectives on understanding the structure and functions of the brain.<br />
<br />
由亨利 · 马克拉姆领导的蓝脑计划旨在建立小鼠大脑的生物细节数字重建和模拟。蓝色大脑项目已经在啮齿动物大脑的电子模型中创建，同时试图尽可能多地复制其生物学细节。基于超级计算机的模拟为理解大脑的结构和功能提供了新的视角。<br />
<br />
The European Union funded a series of projects at the University of Heidelberg, which led to the development of <br />
[[BrainScaleS]] (brain-inspired multiscale computation in neuromorphic hybrid systems), a hybrid analog [[neuromorphic]] supercomputer located at Heidelberg University, Germany. It was developed as part of the [[Human Brain Project]] neuromorphic computing platform and is the complement to the [[SpiNNaker]] supercomputer (which is based on digital technology). The architecture used in BrainScaleS mimics biological neurons and their connections on a physical level; additionally, since the components are made of silicon, these model neurons operate on average 864 times (24 hours of real time is 100 seconds in the machine simulation) that of their biological counterparts.<ref>{{Cite web|date=2016-03-21|title=Beyond von Neumann, Neuromorphic Computing Steadily Advances|url=https://www.hpcwire.com/2016/03/21/lacking-breakthrough-neuromorphic-computing-steadily-advance/|access-date=2021-10-08|website=HPCwire|language=en-US}}</ref><br />
<br />
The European Union funded a series of projects at the University of Heidelberg, which led to the development of <br />
BrainScaleS (brain-inspired multiscale computation in neuromorphic hybrid systems), a hybrid analog neuromorphic supercomputer located at Heidelberg University, Germany. It was developed as part of the Human Brain Project neuromorphic computing platform and is the complement to the SpiNNaker supercomputer (which is based on digital technology). The architecture used in BrainScaleS mimics biological neurons and their connections on a physical level; additionally, since the components are made of silicon, these model neurons operate on average 864 times (24 hours of real time is 100 seconds in the machine simulation) that of their biological counterparts.<br />
<br />
欧盟资助了海德堡大学的一系列项目，这些项目导致了 BrainScaleS (神经形态混合系统中受大脑启发的多尺度计算)的发展，这是一台位于德国海德堡大学的混合模拟神经形态超级计算机。它是作为人脑计划神经形态计算平台的一部分而开发的，是 SpiNNaker 超级计算机(基于数字技术)的补充。大脑尺度中使用的体系结构模拟了生物神经元及其在物理层面上的连接; 此外，由于这些组件是由硅制成的，这些模型神经元平均运行864倍(在机器模拟中，24小时的实时时间是100秒)。<br />
<br />
===Neuromorphic sensors===<br />
===神经形态传感器===<br />
The concept of neuromorphic systems can be extended to sensors (not just to computation). An example of this applied to detecting [[light]] is the [[retinomorphic sensor]] or, when employed in an array, the [[event camera]].<br />
<br />
The concept of neuromorphic systems can be extended to sensors (not just to computation). An example of this applied to detecting light is the retinomorphic sensor or, when employed in an array, the event camera.<br />
<br />
神经形态系统的概念可以扩展到传感器(而不仅仅是计算)。用于检测光线的一个例子是视网膜变形传感器，或者在阵列中使用的事件摄像机。<br />
<br />
==Ethical considerations ==<br />
==伦理问题==<br />
While the interdisciplinary concept of neuromorphic engineering is relatively new, many of the same ethical considerations apply to neuromorphic systems as apply to [[human-like machines]] and [[artificial intelligence]] in general. However, the fact that neuromorphic systems are designed to mimic a [[human brain]] gives rise to unique ethical questions surrounding their usage.<br />
<br />
While the interdisciplinary concept of neuromorphic engineering is relatively new, many of the same ethical considerations apply to neuromorphic systems as apply to human-like machines and artificial intelligence in general. However, the fact that neuromorphic systems are designed to mimic a human brain gives rise to unique ethical questions surrounding their usage.<br />
<br />
虽然神经形态工程的跨学科概念相对较新，但许多同样的伦理考虑适用于神经形态系统，就像适用于类人机器和一般人工智能一样。然而，神经形态系统是为了模仿人类大脑而设计的这一事实引起了围绕其使用的独特的伦理问题。<br />
<br />
However, the practical debate is that neuromorphic hardware as well as artificial "neural networks" are immensely simplified models of how the brain operates or processes information at a much lower [[complex system|complexity in terms of size and functional technology]] and a much more regular structure in terms of [[brain connectivity|connectivity]]. Comparing [[neuromorphic chip]]s to the brain is a very crude comparison similar to comparing a plane to a bird just because they both have wings and a tail. The fact is that neural cognitive systems are many orders of magnitude more [[Energy efficiency (physics)|energy-]] and compute-efficient than current state-of-the-art AI and neuromorphic engineering is an attempt to narrow this gap by inspiring from the brain's mechanism just like many engineering designs have [[bioengineering|bio-inspired features]].<br />
<br />
However, the practical debate is that neuromorphic hardware as well as artificial "neural networks" are immensely simplified models of how the brain operates or processes information at a much lower complexity in terms of size and functional technology and a much more regular structure in terms of connectivity. Comparing neuromorphic chips to the brain is a very crude comparison similar to comparing a plane to a bird just because they both have wings and a tail. The fact is that neural cognitive systems are many orders of magnitude more energy- and compute-efficient than current state-of-the-art AI and neuromorphic engineering is an attempt to narrow this gap by inspiring from the brain's mechanism just like many engineering designs have bio-inspired features.<br />
<br />
然而，实际上的争论是，神经形态硬件和人工“神经网络”是大脑如何运作或处理信息的极其简化的模型，在大小和功能技术方面的复杂性要低得多，在连接方面的结构则更加规则。将神经形态芯片与大脑进行比较是一种非常粗糙的比较，类似于仅仅因为一架飞机有翅膀和一条尾巴就将它与一只鸟进行比较。事实上，神经认知系统比当前最先进的人工智能具有更多的能量和计算效率，而神经形态工程是一种通过从大脑机制中激发灵感来缩小这种差距的尝试，就像许多工程设计具有生物启发的特征一样。数量级。<br />
<br />
===Democratic concerns===<br />
===公众担忧===<br />
Significant ethical limitations may be placed on neuromorphic engineering due to public perception.<ref>{{Cite report|url=https://ai100.stanford.edu/sites/g/files/sbiybj9861/f/ai_100_report_0831fnl.pdf|title=Artificial Intelligence and Life in 2030|author=2015 Study Panel|date=September 2016|work=One Hundred Year Study on Artificial Intelligence (AI100)|publisher=Stanford University}}</ref> Special [[Eurobarometer]] 382: Public Attitudes Towards Robots, a survey conducted by the European Commission, found that 60% of [[European Union]] citizens wanted a ban of robots in the care of children, the elderly, or the disabled. Furthermore, 34% were in favor of a ban on robots in education, 27% in healthcare, and 20% in leisure. The European Commission classifies these areas as notably “human.” The report cites increased public concern with robots that are able to mimic or replicate human functions. Neuromorphic engineering, by definition, is designed to replicate the function of the human brain.<ref name=":1">{{Cite web|url=http://ec.europa.eu/commfrontoffice/publicopinion/archives/ebs/ebs_382_en.pdf|title=Special Eurobarometer 382: Public Attitudes Towards Robots|last=European Commission|date=September 2012|website=European Commission}}</ref><br />
<br />
Significant ethical limitations may be placed on neuromorphic engineering due to public perception. Special Eurobarometer 382: Public Attitudes Towards Robots, a survey conducted by the European Commission, found that 60% of European Union citizens wanted a ban of robots in the care of children, the elderly, or the disabled. Furthermore, 34% were in favor of a ban on robots in education, 27% in healthcare, and 20% in leisure. The European Commission classifies these areas as notably “human.” The report cites increased public concern with robots that are able to mimic or replicate human functions. Neuromorphic engineering, by definition, is designed to replicate the function of the human brain.<br />
<br />
由于公众的认知，神经形态工程学可能会受到严重的伦理限制。欧盟委员会进行的一项调查发现，60% 的欧盟公民希望禁止照顾儿童、老人或残疾人的机器人。此外，34% 的人支持禁止机器人用于教育，27% 的人支持医疗保健，20% 的人支持休闲。欧盟委员会将这些地区明显归类为“人类”报告指出，公众越来越关注能够模仿或复制人类功能的机器人。神经形态工程，顾名思义，是为了复制人脑的功能而设计的。<br />
<br />
The democratic concerns surrounding neuromorphic engineering are likely to become even more profound in the future. The European Commission found that EU citizens between the ages of 15 and 24 are more likely to think of robots as human-like (as opposed to instrument-like) than EU citizens over the age of 55. When presented an image of a robot that had been defined as human-like, 75% of EU citizens aged 15–24 said it corresponded with the idea they had of robots while only 57% of EU citizens over the age of 55 responded the same way. The human-like nature of neuromorphic systems, therefore, could place them in the categories of robots many EU citizens would like to see banned in the future.<ref name=":1" /><br />
<br />
The democratic concerns surrounding neuromorphic engineering are likely to become even more profound in the future. The European Commission found that EU citizens between the ages of 15 and 24 are more likely to think of robots as human-like (as opposed to instrument-like) than EU citizens over the age of 55. When presented an image of a robot that had been defined as human-like, 75% of EU citizens aged 15–24 said it corresponded with the idea they had of robots while only 57% of EU citizens over the age of 55 responded the same way. The human-like nature of neuromorphic systems, therefore, could place them in the categories of robots many EU citizens would like to see banned in the future.<br />
<br />
围绕神经形态工程的民主关注可能在未来变得更加深刻。欧盟委员会(European Commission)发现，15至24岁的欧盟公民比55岁以上的欧盟公民更有可能认为机器人像人(而不是像仪器)。当看到一张被定义为“类人”的机器人图片时，年龄在15岁至24岁之间的欧盟公民中有75% 的人表示，这与他们对机器人的想法相符，而55岁以上的欧盟公民中只有57% 的人有同样的反应。因此，类似人类的神经形态系统，可以把它们归入许多欧盟公民希望在未来禁止使用的机器人类别。<br />
<br />
===Personhood===<br />
===人格问题===<br />
As neuromorphic systems have become increasingly advanced, some scholars{{who|date=August 2021}} have advocated for granting [[personhood]] rights to these systems. If the brain is what grants humans their personhood, to what extent does a neuromorphic system have to mimic the human brain to be granted personhood rights? Critics of technology development in the [[Human Brain Project]], which aims to advance brain-inspired computing, have argued that advancement in neuromorphic computing could lead to machine consciousness or personhood.<ref>{{Cite journal|last=Aicardi|first=Christine|date=September 2018|title=Accompanying technology development in the Human Brain Project: From foresight to ethics management|journal=Futures|volume=102|pages=114–124|doi=10.1016/j.futures.2018.01.005|doi-access=free}}</ref> If these systems are to be treated as people, critics argue, then many tasks humans perform using neuromorphic systems, including the act of termination of neuromorphic systems, may be morally impermissible as these acts would violate the autonomy of the neuromorphic systems.<ref>{{Cite journal|last=Lim|first=Daniel|date=2014-06-01|title=Brain simulation and personhood: a concern with the Human Brain Project|journal=Ethics and Information Technology|language=en|volume=16|issue=2|pages=77–89|doi=10.1007/s10676-013-9330-5|s2cid=17415814|issn=1572-8439}}</ref><br />
<br />
As neuromorphic systems have become increasingly advanced, some scholars have advocated for granting personhood rights to these systems. If the brain is what grants humans their personhood, to what extent does a neuromorphic system have to mimic the human brain to be granted personhood rights? Critics of technology development in the Human Brain Project, which aims to advance brain-inspired computing, have argued that advancement in neuromorphic computing could lead to machine consciousness or personhood. If these systems are to be treated as people, critics argue, then many tasks humans perform using neuromorphic systems, including the act of termination of neuromorphic systems, may be morally impermissible as these acts would violate the autonomy of the neuromorphic systems.<br />
<br />
随着神经形态系统的日益发展，一些学者主张赋予这些系统人格权。如果是大脑赋予了人类人格，那么神经形态系统在多大程度上必须模仿人类大脑才能被赋予人格权利？“人脑计划”旨在推进以大脑为灵感的计算机技术发展，该计划的批评者认为，神经形态计算机技术的进步可能导致机器意识或人格的形成。批评家认为，如果这些系统被当作人来对待，那么人类使用神经形态系统执行的许多任务，包括终止神经形态系统的行为，可能在道德上是不允许的，因为这些行为将违反神经形态系统的自主性。<br />
<br />
==Dual use (military applications)==<br />
==军民两用技术==<br />
The [[Joint Artificial Intelligence Center]], a branch of the U.S. military, is a center dedicated to the procurement and implementation of AI software and neuromorphic hardware for combat use. Specific applications include smart headsets/goggles and robots. JAIC intends to rely heavily on neuromorphic technology to connect "every fighter every shooter" within a network of neuromorphic-enabled units.<br />
<br />
The Joint Artificial Intelligence Center, a branch of the U.S. military, is a center dedicated to the procurement and implementation of AI software and neuromorphic hardware for combat use. Specific applications include smart headsets/goggles and robots. JAIC intends to rely heavily on neuromorphic technology to connect "every fighter every shooter" within a network of neuromorphic-enabled units.<br />
<br />
军民两用联合人工智能中心是美国军队的一个分支，专门从事采购和实施用于战斗的人工智能软件和神经形态硬件。具体应用包括智能耳机/护目镜和机器人。JAIC 打算严重依赖神经形态技术来连接神经形态单位网络中的“每个战士每个射手”。<br />
<br />
==Legal considerations==<br />
==法律问题==<br />
Skeptics have argued that there is no way to apply the electronic personhood, the concept of personhood that would apply to neuromorphic technology, legally. In a letter signed by 285 experts in law, robotics, medicine, and ethics opposing a European Commission proposal to recognize “smart robots” as legal persons, the authors write, “A legal status for a robot can’t derive from the [[Natural person|Natural Person]] model, since the robot would then hold [[human rights]], such as the right to dignity, the right to its integrity, the right to remuneration or the right to citizenship, thus directly confronting the Human rights. This would be in contradiction with the [[Charter of Fundamental Rights of the European Union]] and the [[Convention for the Protection of Human Rights and Fundamental Freedoms]].”<ref>{{Cite web|url=http://www.robotics-openletter.eu/|title=Robotics Openletter {{!}} Open letter to the European Commission|language=fr-FR|access-date=2019-05-10}}</ref><br />
<br />
Skeptics have argued that there is no way to apply the electronic personhood, the concept of personhood that would apply to neuromorphic technology, legally. In a letter signed by 285 experts in law, robotics, medicine, and ethics opposing a European Commission proposal to recognize “smart robots” as legal persons, the authors write, “A legal status for a robot can’t derive from the Natural Person model, since the robot would then hold human rights, such as the right to dignity, the right to its integrity, the right to remuneration or the right to citizenship, thus directly confronting the Human rights. This would be in contradiction with the Charter of Fundamental Rights of the European Union and the Convention for the Protection of Human Rights and Fundamental Freedoms.”<br />
<br />
怀疑论者认为，在法律上没有办法应用电子人格，这个人格概念将适用于神经形态技术。在一封由285名法律、机器人、医学和伦理学专家签名的信中，作者们反对欧盟委员会承认“智能机器人”为法人的提议，他们写道，“机器人的法律地位不能从自然人模型中推导出来，因为机器人将拥有人权，如尊严权、完整权、报酬权或公民权，从而直接面对人权。这将有悖于《欧洲联盟基本权利宪章和《保护人权和基本自由公约》。”<br />
<br />
===Ownership and property rights===<br />
===所有权及财产权问题===<br />
There is significant legal debate around property rights and artificial intelligence. In ''Acohs Pty Ltd v. Ucorp Pty Ltd'', Justice Christopher Jessup of the [[Federal Court of Australia]] found that the [[source code]] for [[Material safety data sheets|Material Safety Data Sheets]] could not be [[Copyright law of Australia|copyrighted]] as it was generated by a [[software interface]] rather than a human author.<ref>{{Cite web|url=http://www.lavan.com.au/advice/intellectual_property_technology/copyright_in_source_code_and_digital_products|title=Copyright in source code and digital products|last=Lavan|website=Lavan|language=en|access-date=2019-05-10}}</ref> The same question may apply to neuromorphic systems: if a neuromorphic system successfully mimics a human brain and produces a piece of original work, who, if anyone, should be able to claim ownership of the work?<ref>{{cite journal |last1=Eshraghian|first1=Jason K. |title=Human Ownership of Artificial Creativity |journal=Nature Machine Intelligence |date=9 March 2020 |volume=2 |pages=157–160 |doi=10.1038/s42256-020-0161-x}}</ref><br />
<br />
There is significant legal debate around property rights and artificial intelligence. In Acohs Pty Ltd v. Ucorp Pty Ltd, Justice Christopher Jessup of the Federal Court of Australia found that the source code for Material Safety Data Sheets could not be copyrighted as it was generated by a software interface rather than a human author. The same question may apply to neuromorphic systems: if a neuromorphic system successfully mimics a human brain and produces a piece of original work, who, if anyone, should be able to claim ownership of the work?<br />
<br />
围绕财产权和人工智能有着重大的法律争论。在 Acohs Pty Ltd 诉 Ucorp Pty Ltd 一案中，澳大利亚联邦法院的克里斯托弗 · 杰瑟普法官发现，材料安全数据表的源代码不能受版权保护，因为它是由软件界面而不是人工作者生成的。同样的问题可能也适用于神经形态系统: 如果一个神经形态系统成功地模仿了人类的大脑并产生了一部原创作品，那么谁，如果有人，应该声称拥有这部作品的所有权？<br />
<br />
== Neuromemristive systems==<br />
==神经记忆电阻系统==<br />
Neuromemristive systems are a subclass of neuromorphic computing systems that focus on the use of [[memristors]] to implement [[neuroplasticity]]. While neuromorphic engineering focuses on mimicking biological behavior, neuromemristive systems focus on abstraction.<ref>{{Cite web|url=https://digitalops.sandia.gov/Mediasite/Play/a10cf6ceb55d47608bb8326dd00e46611d|title=002.08 N.I.C.E. Workshop 2014: Towards Intelligent Computing with Neuromemristive Circuits and Systems - Feb. 2014|website=digitalops.sandia.gov|access-date=2019-08-26}}</ref> For example, a neuromemristive system may replace the details of a [[Cerebral cortex|cortical]] microcircuit's behavior with an abstract neural network model.<ref>C. Merkel and D. Kudithipudi, "Neuromemristive extreme learning machines for pattern classification," ISVLSI, 2014.</ref><br />
<br />
Neuromemristive systems are a subclass of neuromorphic computing systems that focus on the use of memristors to implement neuroplasticity. While neuromorphic engineering focuses on mimicking biological behavior, neuromemristive systems focus on abstraction. For example, a neuromemristive system may replace the details of a cortical microcircuit's behavior with an abstract neural network model.C. Merkel and D. Kudithipudi, "Neuromemristive extreme learning machines for pattern classification," ISVLSI, 2014.<br />
<br />
神经记忆电阻系统是神经形态计算系统的一个亚类，主要研究利用记忆电阻器实现神经可塑性。神经形态工程的重点是模拟生物行为，而神经记忆电阻系统的重点是提取。例如，一个神经记忆系统可以用一个抽象的神经网络模型取代皮层微电路的行为细节。默克尔和 d. Kudithipudi，“用于模式分类的神经记忆极端学习机器”，ISVLSI，2014。<br />
<br />
There exist several neuron inspired threshold logic functions<ref name="Maan 1–13" /> implemented with memristors that have applications in high level [[pattern recognition]] applications. Some of the applications reported recently include [[speech recognition]],<ref>{{Cite journal|title = Memristor pattern recogniser: isolated speech word recognition|journal = Electronics Letters|pages = 1370–1372|volume = 51|issue = 17|doi = 10.1049/el.2015.1428|first1 = A.K.|last1 = Maan|first2 = A.P.|last2 = James|first3 = S.|last3 = Dimitrijev|year = 2015|bibcode = 2015ElL....51.1370M|hdl = 10072/140989|s2cid = 61454815|url = https://semanticscholar.org/paper/48d3ab11ec6e213b62f11eedcfb7b7febb058674|hdl-access = free}}</ref> [[face recognition]]<ref>{{Cite journal|title = Memristive Threshold Logic Face Recognition|journal = Procedia Computer Science|date = 2014-01-01|pages = 98–103|volume = 41|series = 5th Annual International Conference on Biologically Inspired Cognitive Architectures, 2014 BICA|doi = 10.1016/j.procs.2014.11.090|first1 = Akshay Kumar|last1 = Maan|first2 = Dinesh S.|last2 = Kumar|first3 = Alex Pappachen|last3 = James|doi-access = free}}</ref> and [[object recognition]].<ref>{{Cite journal|title = Memristive Threshold Logic Circuit Design of Fast Moving Object Detection|journal = IEEE Transactions on Very Large Scale Integration (VLSI) Systems|date = 2015-10-01|issn = 1063-8210|pages = 2337–2341|volume = 23|issue = 10|doi = 10.1109/TVLSI.2014.2359801|first1 = A.K.|last1 = Maan|first2 = D.S.|last2 = Kumar|first3 = S.|last3 = Sugathan|first4 = A.P.|last4 = James|arxiv = 1410.1267|s2cid = 9647290}}</ref> They also find applications in replacing conventional digital logic gates.<ref>{{Cite journal|title = Resistive Threshold Logic|journal = IEEE Transactions on Very Large Scale Integration (VLSI) Systems|date = 2014-01-01|issn = 1063-8210|pages = 190–195|volume = 22|issue = 1|doi = 10.1109/TVLSI.2012.2232946|first1 = A.P.|last1 = James|first2 = L.R.V.J.|last2 = Francis|first3 = D.S.|last3 = Kumar|arxiv = 1308.0090|s2cid = 7357110}}</ref><ref>{{Cite journal|title = Threshold Logic Computing: Memristive-CMOS Circuits for Fast Fourier Transform and Vedic Multiplication|journal = IEEE Transactions on Very Large Scale Integration (VLSI) Systems|date = 2015-11-01|issn = 1063-8210|pages = 2690–2694|volume = 23|issue = 11|doi = 10.1109/TVLSI.2014.2371857|first1 = A.P.|last1 = James|first2 = D.S.|last2 = Kumar|first3 = A.|last3 = Ajayan|arxiv = 1411.5255|s2cid = 6076956}}</ref><br />
<br />
There exist several neuron inspired threshold logic functions implemented with memristors that have applications in high level pattern recognition applications. Some of the applications reported recently include speech recognition, face recognition and object recognition. They also find applications in replacing conventional digital logic gates.<br />
<br />
利用记忆电阻器实现的神经元启发阈值逻辑函数在高级模式识别中有着广泛的应用。最近报道的一些应用包括语音识别、人脸识别和物体识别。它们还可以用来取代传统的数字逻辑门。<br />
<br />
For ideal passive memristive circuits there is an exact equation (Caravelli-Traversa-[[Di Ventra]] equation) for the internal memory of the circuit:<ref>{{cite journal |last=Caravelli |display-authors=etal|arxiv=1608.08651 |title=The complex dynamics of memristive circuits: analytical results and universal slow relaxation |year=2017 |doi=10.1103/PhysRevE.95.022140 |pmid= 28297937 |volume=95 |issue= 2 |pages= 022140 |journal=Physical Review E|bibcode=2017PhRvE..95b2140C |s2cid=6758362}}</ref><br />
<br />
For ideal passive memristive circuits there is an exact equation (Caravelli-Traversa-Di Ventra equation) for the internal memory of the circuit:<br />
<br />
对于理想的无源记忆电路，电路的内部记忆有一个精确的方程(Caravelli-Traversa-Di Ventra 方程) :<br />
<br />
:<math> \frac{d}{dt} \vec{W} = \alpha \vec{W}-\frac{1}{\beta} (I+\xi \Omega W)^{-1} \Omega \vec S </math><br />
<br />
: \frac{d}{dt} \vec{W} = \alpha \vec{W}-\frac{1}{\beta} (I+\xi \Omega W)^{-1} \Omega \vec S<br />
<br />
:\frac{d}{dt} \vec{W} = \alpha \vec{W}-\frac{1}{\beta} (I+\xi \Omega W)^{-1} \Omega \vec S<br />
<br />
as a function of the properties of the physical memristive network and the external sources. In the equation above, <math>\alpha</math> is the "forgetting" time scale constant, <math> \xi=r-1</math> and <math>r=\frac{R_\text{off}}{R_\text{on}}</math> is the ratio of ''off'' and ''on'' values of the limit resistances of the memristors, <math> \vec S </math> is the vector of the sources of the circuit and <math>\Omega</math> is a projector on the fundamental loops of the circuit. The constant <math>\beta</math> has the dimension of a voltage and is associated to the properties of the [[memristor]]; its physical origin is the charge mobility in the conductor. The diagonal matrix and vector <math>W=\operatorname{diag}(\vec W)</math> and <math>\vec W</math> respectively, are instead the internal value of the memristors, with values between 0 and 1. This equation thus requires adding extra constraints on the memory values in order to be reliable.<br />
<br />
<nowiki>as a function of the properties of the physical memristive network and the external sources. In the equation above, \alpha is the "forgetting" time scale constant, \xi=r-1 and r=\frac{R_\text{off}}{R_\text{on}} is the ratio of off and on values of the limit resistances of the memristors, \vec S is the vector of the sources of the circuit and \Omega is a projector on the fundamental loops of the circuit. The constant \beta has the dimension of a voltage and is associated to the properties of the memristor; its physical origin is the charge mobility in the conductor. The diagonal matrix and vector W=\operatorname{diag}(\vec W) and \vec W respectively, are instead the internal value of the memristors, with values between 0 and 1. This equation thus requires adding extra constraints on the memory values in order to be reliable.</nowiki><br />
<br />
<nowiki>作为物理记忆网络和外部源的性质的函数。在上述方程中，α 是“遗忘”时间尺度常数，xi = r-1，r = frac { r _ text { off }{ on }{ r _ text { on }}是记忆电阻器极限电阻的开关和开关值之比，vec s 是电路源的矢量，Omega 是电路基本环路的投影仪。常数 β 具有电压的尺寸，与记忆电阻器的特性有关; 它的物理起源是导体中的电荷迁移率。对角矩阵和向量 w = 操作者名{ diag }(vec w)和 vec w 分别是记忆电阻器的内值，值在0到1之间。因此，这个等式需要在内存值上添加额外的约束，以保证可靠性。</nowiki><br />
<br />
==See also==<br />
==相关词条==<br />
{{Columns-list|colwidth=18em|<br />
* [[AI accelerator (computer hardware)|AI accelerator]]<br />
* [[Artificial brain]]<br />
* [[Biomorphic]]<br />
* [[Cognitive computer]]<br />
* [[Computation and Neural Systems]]<br />
* [[Differentiable programming]]<br />
* [[Event camera]]<br />
* [[Neurorobotics]]<br />
* [[Optical flow sensor]]<br />
* [[Physical neural network]]<br />
* [[SpiNNaker]]<br />
* [[SyNAPSE]]<br />
* [[Retinomorphic sensor]]<br />
* [[Vision chip]]<br />
* [[Vision processing unit]]<br />
* [[Zeroth (software)]]<br />
* [[Hardware for artificial intelligence]]<br />
}}<br />
<br />
{{Portal bar|Electronics}}<br />
<br />
==References==<br />
{{Reflist|40em}}<br />
<br />
==External links==<br />
<!--======================== {{No more links}} ============================<br />
| PLEASE BE CAUTIOUS IN ADDING MORE LINKS TO THIS ARTICLE. Wikipedia |<br />
| is not a collection of links nor should it be used for advertising. |<br />
| |<br />
| Excessive or inappropriate links WILL BE DELETED. |<br />
| See [[Wikipedia:External links]] & [[Wikipedia:Spam]] for details. |<br />
| |<br />
| If there are already plentiful links, please propose additions or |<br />
| replacements on this article's discussion page, or submit your link |<br />
| to the relevant category at the Open Directory Project (dmoz.org) |<br />
| and link back to that category using the {{dmoz}} template. |<br />
======================={{No more links}}=============================--><br />
{{Commons category|Neuromorphic Engineering}}<br />
*[https://web.archive.org/web/20150727034331/http://ine-web.org/workshops/workshops-overview Telluride Neuromorphic Engineering Workshop]<br />
*[https://archive.today/20130115190057/http://capocaccia.ethz.ch/ CapoCaccia Cognitive Neuromorphic Engineering Workshop]<br />
*[https://web.archive.org/web/20190716132350/http://www.ine-web.org/ Institute of Neuromorphic Engineering]<br />
*[http://www.ine-news.org/ INE news site].<br />
*[http://www.frontiersin.org/neuromorphic_engineering Frontiers in Neuromorphic Engineering Journal]<br />
*[http://www.cns.caltech.edu/ Computation and Neural Systems] department at the [[California Institute of Technology]].<br />
*[http://www.humanbrainproject.eu/ Human Brain Project official site]<br />
*[https://www.the-scientist.com/features/building-a-silicon-brain-65738 Building a Silicon Brain:] Computer chips based on biological neurons may help simulate larger and more-complex brain models. May 1, 2019. SANDEEP RAVINDRAN<br />
<br />
<br />
*Telluride Neuromorphic Engineering Workshop<br />
*CapoCaccia Cognitive Neuromorphic Engineering Workshop<br />
*Institute of Neuromorphic Engineering<br />
*INE news site.<br />
*Frontiers in Neuromorphic Engineering Journal<br />
*Computation and Neural Systems department at the California Institute of Technology.<br />
*Human Brain Project official site<br />
*Building a Silicon Brain: Computer chips based on biological neurons may help simulate larger and more-complex brain models. May 1, 2019. SANDEEP RAVINDRAN<br />
<br />
=<nowiki>外部链接</nowiki>= <br />
*碲化物神经形态工程工作室<br />
*CapoCaccia 认知神经形态工程工作室<br />
*神经形态工程研究所<br />
*INE 新闻站点。<br />
*《神经形态工程学前沿》<br />
*加州理工学院计算与神经系统系。<br />
*人脑项目官方网站<br />
*打造硅脑: 基于生物神经元的计算机芯片可能有助于模拟更大、更复杂的大脑模型。2019年5月1日。SANDEEP RAVINDRAN<br />
<br />
{{Differentiable computing}}<br />
{{Authority control}}<br />
<br />
[[Category:Electrical engineering]]<br />
[[Category:Neuroscience]]<br />
[[分类:AI accelerators|*]]<br />
[[Category:Artificial intelligence]]<br />
[[Category:Robotics]]<br />
<br />
Category:Electrical engineering<br />
Category:Neuroscience<br />
*<br />
Category:Artificial intelligence<br />
Category:Robotics<br />
<br />
类别: 电气工程类别: 神经科学 * 类别: 人工智能类别: 机器人<br />
<br />
<noinclude><br />
<br />
<small>This page was moved from [[wikipedia:en:Neuromorphic engineering]]. Its edit history can be viewed at [[神经形态计算/edithistory]]</small></noinclude><br />
<br />
[[Category:待整理页面]]<br />
<br />
<references /></div>Shenky20https://wiki.swarma.org/index.php?title=%E7%A5%9E%E7%BB%8F%E5%BD%A2%E6%80%81%E8%AE%A1%E7%AE%97&diff=30176神经形态计算2022-04-13T07:56:49Z<p>Shenky20：</p>
<hr />
<div><br />
此词条由神经动力学模型读书会词条梳理志愿者Shenky20翻译审校，未经专家审校，带来阅读不便，请见谅。<br />
<br />
{{Use American English|date = January 2019}}<br />
{{Use mdy dates|date = January 2019}}<br />
<br />
'''Neuromorphic engineering''', also known as '''neuromorphic computing''',<ref name=":3">{{Cite journal | doi = 10.1145/2601069| title = Neuromorphic computing gets ready for the (really) big time| journal = [[Communications of the ACM]]| volume = 57| issue = 6| year = 2014| last1 = Monroe | first1 = D. | pages = 13–15| s2cid = 20051102}}</ref><ref name=":4">{{Cite journal | doi = 10.1088/0957-4484/21/17/175202| pmid = 20368686| title = Nanotube devices based crossbar architecture: Toward neuromorphic computing| journal = Nanotechnology| volume = 21| issue = 17| pages = 175202| year = 2010| last1 = Zhao | first1 = W. S. | last2 = Agnus | first2 = G. | last3 = Derycke | first3 = V. | last4 = Filoramo | first4 = A. | last5 = Bourgoin | first5 = J. -P. | last6 = Gamrat | first6 = C. | bibcode = 2010Nanot..21q5202Z| url = https://zenodo.org/record/3428659}}</ref><ref name="humanbrainproject">{{YouTube|id=6RoiZ90mGfw|title=The Human Brain Project SP 9: Neuromorphic Computing Platform}}</ref> is the use of [[very-large-scale integration]] (VLSI) systems containing electronic [[analog circuit]]s to mimic neuro-biological architectures present in the nervous system. A neuromorphic computer/chip is any device that uses physical artificial neurons (made from silicon) to do computations.<ref name=":5">{{cite journal|last1=Mead|first1=Carver|title=Neuromorphic electronic systems|journal=Proceedings of the IEEE|date=1990|volume=78|issue=10|pages=1629–1636|doi=10.1109/5.58356|url=https://authors.library.caltech.edu/53090/1/00058356.pdf}}</ref><ref name=":2" /> In recent times, the term ''neuromorphic'' has been used to describe analog, digital, mixed-mode analog/digital VLSI, and software systems that implement models of [[neural system]]s (for [[perception]], [[motor control]], or [[multisensory integration]]). The implementation of neuromorphic computing on the hardware level can be realized by oxide-based [[memristor]]s,<ref name="Maan 1–13">{{Cite journal|last1=Maan|first1=A. K.|last2=Jayadevi|first2=D. A.|last3=James|first3=A. P.|date=2016-01-01|title=A Survey of Memristive Threshold Logic Circuits|journal=IEEE Transactions on Neural Networks and Learning Systems|volume=PP|issue=99|pages=1734–1746|doi=10.1109/TNNLS.2016.2547842|pmid=27164608|issn=2162-237X|arxiv=1604.07121|bibcode=2016arXiv160407121M|s2cid=1798273}}</ref> spintronic memories, threshold switches, and [[transistor]]s.<ref name=":6">{{Cite journal|title = Mott Memory and Neuromorphic Devices|journal = Proceedings of the IEEE|date = 2015-08-01|issn = 0018-9219|pages = 1289–1310|volume = 103|issue = 8|doi = 10.1109/JPROC.2015.2431914|first1 = You|last1 = Zhou|first2 = S.|last2 = Ramanathan|s2cid = 11347598|url = https://zenodo.org/record/895565}}</ref><ref name=":2">{{Cite document|title=Neuromorphic Circuits With Neural Modulation Enhancing the Information Content of Neural Signaling {{!}} International Conference on Neuromorphic Systems 2020|language=EN|doi=10.1145/3407197.3407204|s2cid=220794387}}</ref> Training software-based neuromorphic systems of spiking neural networks can be achieved using error backpropagation, e.g., using [[Python (programming language)|Python]] based frameworks such as snnTorch,<ref name=":7">{{cite journal |last1=Eshraghian|first1=Jason K.|last2=Ward|first2=Max|last3=Neftci |first3=Emre|last4=Wang|first4=Xinxin|last5=Lenz|first5=Gregor|last6=Dwivedi|first6=Girish|last7=Bennamoun|first7=Mohammed|last8=Jeong|first8=Doo Seok|last9=Lu|first9=Wei D.|title=Training Spiking Neural Networks Using Lessons from Deep Learning |date=1 October 2021 |arxiv=2109.12894 }}</ref> or using canonical learning rules from the biological learning literature, e.g., using BindsNet.<ref name=":8">{{Cite web | url=https://github.com/Hananel-Hazan/bindsnet | title=Hananel-Hazan/bindsnet: Simulation of spiking neural networks (SNNs) using PyTorch.| date=31 March 2020}}</ref><br />
<br />
<br />
'''<font color="#ff8000">神经形态工程Neuromorphic engineering</font>'''(也称为'''<font color="#ff8000">神经形态计算Neuromorphic computing</font>''')<ref name=":3" /><ref name=":4" /><ref name="humanbrainproject" />是指使用包含电子'''<font color="#ff8000">模拟电路Analog circuit</font>'''的'''<font color="#ff8000">超大规模集成电路Very-large-scale integration</font>'''系统来模拟神经系统中生理结构的研究方法。神经形态计算机或神经形态芯片包括任何使用由硅制成的人造神经元进行计算的设备。<ref name=":5" /><ref name=":2" />近年来，神经形态学（neuromorphic）这个术语被用来描述能够实现'''<font color="#ff8000">神经系统Neural system</font>'''模型功能(如'''<font color="#ff8000">感知Perception</font>'''、'''<font color="#ff8000">运动控制Motor control</font>'''，'''<font color="#ff8000">多感官整合Multisensory integration</font>'''等)的模拟、数字、模拟/数字混合模式超大规模集成电路和软件系统。神经形态计算的硬件实现可以通过基于氧化物的'''<font color="#ff8000">记忆电阻器Memristor</font>'''、<ref name="Maan 1–13" />自旋电子存储器、阈值开关和'''<font color="#ff8000">晶体管Transistor</font>'''来实现。<ref name=":6" /><ref name=":2" />对基于软件的脉冲神经网络系统的训练可以通过误差反向传播机制来实现，例如，使用snnTorch等基于Python的框架，<ref name=":7" />或使用BindsNet等典型的受生物启发的学习模式。<ref name=":8" /><br />
<br />
A key aspect of neuromorphic engineering is understanding how the morphology of individual neurons, circuits, applications, and overall architectures creates desirable computations, affects how information is represented, influences robustness to damage, incorporates learning and development, adapts to local change (plasticity), and facilitates evolutionary change.<br />
<br />
<br />
神经形态工程领域的一个关键问题，就是理解'''<font color="#32CD32">单个神经元形态、神经回路、应用和整体结构</font>'''如何产生理想的计算，如何影响信息的表示和对破坏的鲁棒性，如何整合学习和发展，如何适应局部变化(可塑性) 并促进逐渐发展的变化。<br />
<br />
Neuromorphic engineering is an interdisciplinary subject that takes inspiration from [[biology]], [[physics]], [[mathematics]], [[computer science]], and [[electronic engineering]]<ref name=":2" /> to design artificial neural systems, such as [[Machine vision|vision systems]], head-eye systems, auditory processors, and autonomous robots, whose physical architecture and design principles are based on those of biological nervous systems.<ref name=":9">{{Cite journal | doi = 10.1155/2012/705483| title = Qualitative Functional Decomposition Analysis of Evolved Neuromorphic Flight Controllers| journal = Applied Computational Intelligence and Soft Computing| volume = 2012| pages = 1–21| year = 2012| last1 = Boddhu | first1 = S. K. | last2 = Gallagher | first2 = J. C. | doi-access = free}}</ref> It was developed by [[Carver Mead]]<ref name=":10">{{cite web|title=carver mead website|last1=Mead|first1=Carver|url=http://www.carvermead.caltech.edu/index.html|website=carvermead}}</ref> in the late 1980s.<br />
<br />
<br />
神经形态工程是以'''<font color="#ff8000">生物学Biology</font>'''、'''<font color="#ff8000">物理学Physics</font>'''、'''<font color="#ff8000">数学Mathematics</font>'''、'''<font color="#ff8000">计算机科学Computer science</font>'''和'''<font color="#ff8000">电子工程Electronic engineering<ref name=":2" /></font>'''等学科为基础，设计人工神经系统（如'''<font color="#ff8000">视觉系统Vision systems</font>'''、头眼系统、听觉处理器以及物理结构和设计原则都受启发于生物神经系统的自主机器人）的一门交叉学科。<ref name=":9" />20世纪80年代后期，卡弗·米德极大地推动了神经形态工程领域的发展。<ref name=":10" /><br />
<br />
==Neurological inspiration==<br />
==神经生物学启发==<br />
Neuromorphic engineering is set apart by the inspiration it takes from what we know about the structure and operations of the [[brain]]. Neuromorphic engineering translates what we know about the brain's function into computer systems. Work has mostly focused on replicating the analog nature of [[biological computation]] and the role of [[neuron]]s in [[cognition]].<br />
<br />
Neuromorphic engineering is set apart by the inspiration it takes from what we know about the structure and operations of the brain. Neuromorphic engineering translates what we know about the brain's function into computer systems. Work has mostly focused on replicating the analog nature of biological computation and the role of neurons in cognition.<br />
<br />
神经学灵感工程学是从我们对大脑结构和运作的了解中获得灵感而形成的。神经形态工程学将我们对大脑功能的了解转化为计算机系统。工作主要集中在复制生物计算的模拟特性和神经元在认知中的作用。<br />
<br />
The biological processes of neurons and their [[synapse]]s are dauntingly complex, and thus very difficult to artificially simulate. A key feature of biological brains is that all of the processing in neurons use analog chemical signals. This makes it hard to replicate brains in computers because the current generation of computers is completely digital. However, the characteristics of these parts can be abstracted into mathematical functions that closely capture the essence of the neuron's operations.<br />
<br />
The biological processes of neurons and their synapses are dauntingly complex, and thus very difficult to artificially simulate. A key feature of biological brains is that all of the processing in neurons use analog chemical signals. This makes it hard to replicate brains in computers because the current generation of computers is completely digital. However, the characteristics of these parts can be abstracted into mathematical functions that closely capture the essence of the neuron's operations.<br />
<br />
神经元及其突触的生物学过程极其复杂，难以人工模拟。生物大脑的一个关键特征是，神经元的所有处理过程都使用模拟化学信号。这使得很难在计算机中复制大脑，因为目前的计算机是完全数字化的。然而，这些部分的特征可以抽象为数学函数，这些函数紧密地抓住了神经元运行的本质。<br />
<br />
The goal of neuromorphic computing is not to perfectly mimic the brain and all of its functions, but instead to extract what is known of its structure and operations to be used in a practical computing system. No neuromorphic system will claim nor attempt to reproduce every element of neurons and synapses, but all adhere to the idea that computation is highly [[distributed processing|distributed]] throughout a series of small computing elements analogous to a neuron. While this sentiment is standard, researchers chase this goal with different methods.<ref>{{Cite journal | doi = 10.1088/1741-2560/13/5/051001| title = Large-scale neuromorphic computing systems| journal = Journal of Neural Engineering| volume = 13| pages = 1–15| year = 2016| last1 = Furber | first1 = Steve| issue = 5| pmid = 27529195| bibcode = 2016JNEng..13e1001F| doi-access = free}}</ref><br />
<br />
The goal of neuromorphic computing is not to perfectly mimic the brain and all of its functions, but instead to extract what is known of its structure and operations to be used in a practical computing system. No neuromorphic system will claim nor attempt to reproduce every element of neurons and synapses, but all adhere to the idea that computation is highly distributed throughout a series of small computing elements analogous to a neuron. While this sentiment is standard, researchers chase this goal with different methods.<br />
<br />
神经形态计算的目标不是完美地模拟大脑及其所有功能，而是提取已知的大脑结构和操作，用于实际的计算系统。没有哪个神经形态学系统会声称或试图复制神经元和突触的每一个元素，但所有人都坚持这样的观点，即计算是高度分布在一系列类似于神经元的小型计算元素中的。虽然这种情绪是标准的，但研究人员用不同的方法追求这一目标。<br />
<br />
==Examples==<br />
==案例==<br />
As early as 2006, researchers at [[Georgia Tech]] published a field programmable neural array.<ref>{{Cite book|title = A field programmable neural array|last1 = Farquhar|first1 = Ethan|date = May 2006|journal = IEEE International Symposium on Circuits and Systems|pages = 4114–4117|last2 = Hasler|first2 = Paul.|doi = 10.1109/ISCAS.2006.1693534|isbn = 978-0-7803-9389-9|s2cid = 206966013}}</ref> This chip was the first in a line of increasingly complex arrays of floating gate transistors that allowed programmability of charge on the gates of [[MOSFET]]s to model the channel-ion characteristics of neurons in the brain and was one of the first cases of a silicon programmable array of neurons.<br />
<br />
As early as 2006, researchers at Georgia Tech published a field programmable neural array. This chip was the first in a line of increasingly complex arrays of floating gate transistors that allowed programmability of charge on the gates of MOSFETs to model the channel-ion characteristics of neurons in the brain and was one of the first cases of a silicon programmable array of neurons.<br />
<br />
早在2006年，佐治亚理工学院的研究人员就发表了现场可编程神经阵列。这种芯片是一系列越来越复杂的浮栅晶体管阵列中的第一个，这些晶体管可以在 mosfet 的栅极上编程来模拟大脑中神经元的通道离子特性，这也是硅可编程神经元阵列的首批案例之一。<br />
<br />
In November 2011, a group of [[MIT]] researchers created a computer chip that mimics the analog, ion-based communication in a synapse between two neurons using 400 transistors and standard [[CMOS]] manufacturing techniques.<ref>{{cite web|title=MIT creates "brain chip"|url=http://www.extremetech.com/extreme/105067-mit-creates-brain-chip|access-date=4 December 2012}}</ref><ref name="Neuromorphic silicon paper">{{cite journal|title=Neuromorphic silicon neurons and large-scale neural networks: challenges and opportunities|doi=10.3389/fnins.2011.00108|pmid=21991244|pmc=3181466|volume=5|pages=108|journal=Frontiers in Neuroscience|year=2011|last1=Poon|first1=Chi-Sang|last2=Zhou|first2=Kuan|doi-access=free}}</ref><br />
<br />
In November 2011, a group of MIT researchers created a computer chip that mimics the analog, ion-based communication in a synapse between two neurons using 400 transistors and standard CMOS manufacturing techniques.<br />
<br />
2011年11月，麻省理工学院的一组研究人员发明了一种计算机芯片，该芯片使用400个晶体管和标准的 CMOS 制造技术，在两个神经元之间的突触中模拟模拟基于离子的通讯。<br />
<br />
In June 2012, [[spintronic]] researchers at [[Purdue University]] presented a paper on the design of a neuromorphic chip using [[Spin valve|lateral spin valve]]s and [[memristor]]s. They argue that the architecture works similarly to neurons and can therefore be used to test methods of reproducing the brain's processing. In addition, these chips are significantly more energy-efficient than conventional ones.<ref name="Spin Devices Prop">{{Cite arXiv|title=Proposal For Neuromorphic Hardware Using Spin Devices|eprint=1206.3227|last1=Sharad|first1=Mrigank|last2=Augustine|first2=Charles|last3=Panagopoulos|first3=Georgios|last4=Roy|first4=Kaushik|class=cond-mat.dis-nn|year=2012}}</ref><br />
<br />
In June 2012, spintronic researchers at Purdue University presented a paper on the design of a neuromorphic chip using lateral spin valves and memristors. They argue that the architecture works similarly to neurons and can therefore be used to test methods of reproducing the brain's processing. In addition, these chips are significantly more energy-efficient than conventional ones.<br />
<br />
2012年6月，普渡大学的自旋电子学研究人员发表了一篇关于利用侧向自旋阀和记忆电阻器设计神经形态芯片的论文。他们认为，这种结构的工作原理与神经元相似，因此可以用来测试复制大脑处理过程的方法。此外，这些芯片明显比传统芯片更节能。<br />
<br />
Research at [[HP Labs]] on Mott memristors has shown that while they can be non-[[Volatile memory|volatile]], the volatile behavior exhibited at temperatures significantly below the [[phase transition]] temperature can be exploited to fabricate a [[neuristor]],<ref name=":0" /> a biologically-inspired device that mimics behavior found in neurons.<ref name=":0">{{Cite journal | doi = 10.1038/nmat3510| pmid = 23241533| title = A scalable neuristor built with Mott memristors| journal = Nature Materials| volume = 12| issue = 2| pages = 114–7| year = 2012| last1 = Pickett | first1 = M. D. | last2 = Medeiros-Ribeiro | first2 = G. | last3 = Williams | first3 = R. S. | bibcode = 2013NatMa..12..114P| s2cid = 16271627| url = https://semanticscholar.org/paper/b6ba6f496ace2b947f111059663e76bb60e9efeb}}</ref> In September 2013, they presented models and simulations that show how the spiking behavior of these neuristors can be used to form the components required for a [[Turing machine]].<ref>{{cite journal|doi=10.1088/0957-4484/24/38/384002|title=Phase transitions enable computational universality in neuristor-based cellular automata|author1=Matthew D Pickett|author2=R Stanley Williams|name-list-style=amp|date=September 2013|publisher=IOP Publishing Ltd|journal=Nanotechnology|volume=24|issue=38|pmid=23999059|bibcode=2013Nanot..24L4002P|at=384002}}</ref><br />
<br />
Research at HP Labs on Mott memristors has shown that while they can be non-volatile, the volatile behavior exhibited at temperatures significantly below the phase transition temperature can be exploited to fabricate a neuristor, a biologically-inspired device that mimics behavior found in neurons. In September 2013, they presented models and simulations that show how the spiking behavior of these neuristors can be used to form the components required for a Turing machine.<br />
<br />
惠普实验室在莫特记忆电阻器上的研究表明，尽管它们可以是非挥发性的，但是在相变温度以下显示的挥发性行为可以被用来制造神经元电阻器，这是一种模仿神经元行为的生物设备。2013年9月，他们展示了模型和仿真，展示了这些神经元的尖峰行为是如何被用来形成图灵机所需的元件的。<br />
<br />
[[Neurogrid]], built by ''Brains in Silicon'' at [[Stanford University]],<ref>{{cite journal|last1=Boahen|first1=Kwabena|title=Neurogrid: A Mixed-Analog-Digital Multichip System for Large-Scale Neural Simulations|journal=Proceedings of the IEEE|date=24 April 2014|volume=102|issue=5|pages=699–716|doi=10.1109/JPROC.2014.2313565|s2cid=17176371}}</ref> is an example of hardware designed using neuromorphic engineering principles. The circuit board is composed of 16 custom-designed chips, referred to as NeuroCores. Each NeuroCore's analog circuitry is designed to emulate neural elements for 65536 neurons, maximizing energy efficiency. The emulated neurons are connected using digital circuitry designed to maximize spiking throughput.<ref>{{cite journal|doi=10.1038/503022a|pmid = 24201264|title = Neuroelectronics: Smart connections|journal = Nature|volume = 503|issue = 7474|pages = 22–4|year = 2013|last1 = Waldrop|first1 = M. Mitchell|bibcode = 2013Natur.503...22W|doi-access = free}}</ref><ref>{{cite journal|doi=10.1109/JPROC.2014.2313565|title = Neurogrid: A Mixed-Analog-Digital Multichip System for Large-Scale Neural Simulations|journal = Proceedings of the IEEE|volume = 102|issue = 5|pages = 699–716|year = 2014|last1 = Benjamin|first1 = Ben Varkey|last2 = Peiran Gao|last3 = McQuinn|first3 = Emmett|last4 = Choudhary|first4 = Swadesh|last5 = Chandrasekaran|first5 = Anand R.|last6 = Bussat|first6 = Jean-Marie|last7 = Alvarez-Icaza|first7 = Rodrigo|last8 = Arthur|first8 = John V.|last9 = Merolla|first9 = Paul A.|last10 = Boahen|first10 = Kwabena|s2cid = 17176371}}</ref><br />
<br />
Neurogrid, built by Brains in Silicon at Stanford University, is an example of hardware designed using neuromorphic engineering principles. The circuit board is composed of 16 custom-designed chips, referred to as NeuroCores. Each NeuroCore's analog circuitry is designed to emulate neural elements for 65536 neurons, maximizing energy efficiency. The emulated neurons are connected using digital circuitry designed to maximize spiking throughput.<br />
<br />
神经网格是由斯坦福大学硅谷大脑公司建立的，是一个使用神经形态工程原理设计硬件的例子。该电路板由16个定制设计的芯片组成，称为 NeuroCores。每个 NeuroCore 的模拟电路被设计为模拟65536个神经元的神经元元件，最大限度地提高能量效率。模拟的神经元通过设计的数字电路连接，以最大化脉冲吞吐量。<br />
<br />
A research project with implications for neuromorphic engineering is the [[Human Brain Project]] that is attempting to simulate a complete human brain in a supercomputer using biological data. It is made up of a group of researchers in neuroscience, medicine, and computing.<ref>{{cite web|title=Involved Organizations|url=http://www.humanbrainproject.eu/partners.html|access-date=22 February 2013|url-status=dead|archive-url=https://web.archive.org/web/20130302142627/http://www.humanbrainproject.eu/partners.html|archive-date=2 March 2013}}</ref> [[Henry Markram]], the project's co-director, has stated that the project proposes to establish a foundation to explore and understand the brain and its diseases, and to use that knowledge to build new computing technologies. The three primary goals of the project are to better understand how the pieces of the brain fit and work together, to understand how to objectively diagnose and treat brain diseases, and to use the understanding of the human brain to develop neuromorphic computers. That the simulation of a complete human brain will require a supercomputer a thousand times more powerful than today's encourages the current focus on neuromorphic computers.<ref>{{cite web|title=Human Brain Project|url=http://www.humanbrainproject.eu|access-date=22 February 2013}}</ref> $1.3 billion has been allocated to the project by The [[European Commission]].<ref>{{cite web|title=The Human Brain Project and Recruiting More Cyberwarriors|url=http://www.marketplace.org/topics/tech/human-brain-project-and-recruiting-more-cyberwarriors|access-date=22 February 2013|date=January 29, 2013}}</ref><br />
<br />
A research project with implications for neuromorphic engineering is the Human Brain Project that is attempting to simulate a complete human brain in a supercomputer using biological data. It is made up of a group of researchers in neuroscience, medicine, and computing. Henry Markram, the project's co-director, has stated that the project proposes to establish a foundation to explore and understand the brain and its diseases, and to use that knowledge to build new computing technologies. The three primary goals of the project are to better understand how the pieces of the brain fit and work together, to understand how to objectively diagnose and treat brain diseases, and to use the understanding of the human brain to develop neuromorphic computers. That the simulation of a complete human brain will require a supercomputer a thousand times more powerful than today's encourages the current focus on neuromorphic computers. $1.3 billion has been allocated to the project by The European Commission.<br />
<br />
一个对神经形态工程有影响的研究项目是人脑项目，它试图用生物数据在超级计算机中模拟完整的人脑。它由一群神经科学、医学和计算机科学的研究人员组成。该项目的联合主管亨利•马克拉姆(Henry Markram)表示，该项目提议建立一个基础，以探索和了解大脑及其疾病，并利用这些知识构建新的计算机技术。这个项目的三个主要目标是: 更好地理解大脑的各个部分是如何相互匹配和协同工作的; 理解如何客观地诊断和治疗脑部疾病; 以及利用对人类大脑的理解来开发神经形态计算机。模拟一个完整的人类大脑需要一台比现在强大一千倍的超级计算机，这鼓励了当前对神经形态计算机的关注。欧洲委员会已经拨款13亿美元用于这个项目。<br />
<br />
Other research with implications for neuromorphic engineering involves the [[BRAIN Initiative]]<ref name="economist">[https://www.economist.com/news/science-and-technology/21582495-computers-will-help-people-understand-brains-better-and-understanding-brains Neuromorphic computing: The machine of a new soul], The Economist, 2013-08-03</ref> and the [[TrueNorth]] chip from [[IBM]].<ref>{{cite journal|last1=Modha|first1=Dharmendra|title=A million spiking-neuron integrated circuit with a scalable communication network and interface|journal=Science|date=Aug 2014|volume=345|issue=6197|pages=668–673|doi=10.1126/science.1254642|pmid=25104385|bibcode=2014Sci...345..668M|s2cid=12706847}}</ref> Neuromorphic devices have also been demonstrated using nanocrystals, nanowires, and conducting polymers.<ref>{{Cite web|url=http://jessamynfairfield.com/wp-content/uploads/2017/03/PWMar17Fairfield.pdf|title=Smarter Machines|last=Fairfield|first=Jessamyn|date=March 1, 2017}}</ref><br />
<br />
Other research with implications for neuromorphic engineering involves the BRAIN InitiativeNeuromorphic computing: The machine of a new soul, The Economist, 2013-08-03 and the TrueNorth chip from IBM. Neuromorphic devices have also been demonstrated using nanocrystals, nanowires, and conducting polymers.<br />
<br />
其他与神经形态工程有关的研究还包括 BRAIN initiativity/euromorphic computing: The machine of a new soul，The Economist，2013-08-03 and The TrueNorth chip from IBM。神经形态学设备也已经被证明使用纳米晶体、纳米线和导电聚合物。<br />
<br />
[[Intel]] unveiled its neuromorphic research chip, called “[[Intel Loihi|Loihi]]”, in October 2017. The chip uses an asynchronous [[spiking neural network]] (SNN) to implement adaptive self-modifying event-driven fine-grained parallel computations used to implement learning and inference with high efficiency.<ref>{{cite journal |last1=Davies |first1=Mike |title=Loihi: A Neuromorphic Manycore Processor with On-Chip Learning |journal=IEEE Micro |date=January 16, 2018 |volume=38 |issue=1 |pages=82–99 |display-authors=etal|doi=10.1109/MM.2018.112130359 |s2cid=3608458 }}</ref><ref name="Morris2017">{{cite web |last1=Morris |first1=John |title=Why Intel built a neuromorphic chip |url=https://www.zdnet.com/article/why-intel-built-a-neuromorphic-chip/ |website=ZDNet |access-date=17 August 2018 |language=en}}</ref><br />
<br />
Intel unveiled its neuromorphic research chip, called “Loihi”, in October 2017. The chip uses an asynchronous spiking neural network (SNN) to implement adaptive self-modifying event-driven fine-grained parallel computations used to implement learning and inference with high efficiency.<br />
<br />
2017年10月，英特尔公布了其神经形态研究芯片“ Loihi”。该芯片采用异步脉冲神经网络(SNN)实现自适应自修改事件驱动的细粒度并行计算，实现了高效的学习和推理。<br />
<br />
[[IMEC]], a Belgium-based nanoelectronics research center, demonstrated the world's first self-learning neuromorphic chip. The brain-inspired chip, based on OxRAM technology, has the capability of self-learning and has been demonstrated to have the ability to compose music.<ref>{{cite web |title=Imec demonstrates self-learning neuromorphic chip that composes music |url=https://www.imec-int.com/en/articles/imec-demonstrates-self-learning-neuromorphic-chip-that-composes-music |website=IMEC International |access-date=1 October 2019}}</ref> IMEC released the 30-second tune composed by the prototype. The chip was sequentially loaded with songs in the same time signature and style. The songs were old Belgian and French flute minuets, from which the chip learned the rules at play and then applied them.<ref>{{cite web|last1=Bourzac|first1=Katherine|title=A Neuromorphic Chip That Makes Music|url=https://spectrum.ieee.org/a-neuromorphic-chip-that-makes-music|url-status=live|access-date=1 October 2019|website=IEEE Spectrum|date=May 23, 2017}}</ref><br />
<br />
IMEC, a Belgium-based nanoelectronics research center, demonstrated the world's first self-learning neuromorphic chip. The brain-inspired chip, based on OxRAM technology, has the capability of self-learning and has been demonstrated to have the ability to compose music. IMEC released the 30-second tune composed by the prototype. The chip was sequentially loaded with songs in the same time signature and style. The songs were old Belgian and French flute minuets, from which the chip learned the rules at play and then applied them.<br />
<br />
比利时的一个纳米电子学研究中心 IMEC 展示了世界上第一个自学习神经形态芯片。这种基于 OxRAM 技术的大脑启发芯片具有自学习能力，并已被证明具有创作音乐的能力。IMEC 发布了由原型机谱写的30秒曲调。芯片按顺序加载同时签名和风格的歌曲。歌曲是古老的比利时和法国长笛小步舞曲，筹码从中学习游戏规则，然后应用它们。<br />
<br />
[[Blue Brain Project|The Blue Brain Project]], led by Henry Markram, aims to build biologically detailed digital reconstructions and simulations of the mouse brain. The Blue Brain Project has created in silico models of rodent brains, while attempting to replicate as many details about its biology as possible. The supercomputer-based simulations offer new perspectives on understanding the structure and functions of the brain.<br />
<br />
The Blue Brain Project, led by Henry Markram, aims to build biologically detailed digital reconstructions and simulations of the mouse brain. The Blue Brain Project has created in silico models of rodent brains, while attempting to replicate as many details about its biology as possible. The supercomputer-based simulations offer new perspectives on understanding the structure and functions of the brain.<br />
<br />
由亨利 · 马克拉姆领导的蓝脑计划旨在建立小鼠大脑的生物细节数字重建和模拟。蓝色大脑项目已经在啮齿动物大脑的电子模型中创建，同时试图尽可能多地复制其生物学细节。基于超级计算机的模拟为理解大脑的结构和功能提供了新的视角。<br />
<br />
The European Union funded a series of projects at the University of Heidelberg, which led to the development of <br />
[[BrainScaleS]] (brain-inspired multiscale computation in neuromorphic hybrid systems), a hybrid analog [[neuromorphic]] supercomputer located at Heidelberg University, Germany. It was developed as part of the [[Human Brain Project]] neuromorphic computing platform and is the complement to the [[SpiNNaker]] supercomputer (which is based on digital technology). The architecture used in BrainScaleS mimics biological neurons and their connections on a physical level; additionally, since the components are made of silicon, these model neurons operate on average 864 times (24 hours of real time is 100 seconds in the machine simulation) that of their biological counterparts.<ref>{{Cite web|date=2016-03-21|title=Beyond von Neumann, Neuromorphic Computing Steadily Advances|url=https://www.hpcwire.com/2016/03/21/lacking-breakthrough-neuromorphic-computing-steadily-advance/|access-date=2021-10-08|website=HPCwire|language=en-US}}</ref><br />
<br />
The European Union funded a series of projects at the University of Heidelberg, which led to the development of <br />
BrainScaleS (brain-inspired multiscale computation in neuromorphic hybrid systems), a hybrid analog neuromorphic supercomputer located at Heidelberg University, Germany. It was developed as part of the Human Brain Project neuromorphic computing platform and is the complement to the SpiNNaker supercomputer (which is based on digital technology). The architecture used in BrainScaleS mimics biological neurons and their connections on a physical level; additionally, since the components are made of silicon, these model neurons operate on average 864 times (24 hours of real time is 100 seconds in the machine simulation) that of their biological counterparts.<br />
<br />
欧盟资助了海德堡大学的一系列项目，这些项目导致了 BrainScaleS (神经形态混合系统中受大脑启发的多尺度计算)的发展，这是一台位于德国海德堡大学的混合模拟神经形态超级计算机。它是作为人脑计划神经形态计算平台的一部分而开发的，是 SpiNNaker 超级计算机(基于数字技术)的补充。大脑尺度中使用的体系结构模拟了生物神经元及其在物理层面上的连接; 此外，由于这些组件是由硅制成的，这些模型神经元平均运行864倍(在机器模拟中，24小时的实时时间是100秒)。<br />
<br />
===Neuromorphic sensors===<br />
===神经形态传感器===<br />
The concept of neuromorphic systems can be extended to sensors (not just to computation). An example of this applied to detecting [[light]] is the [[retinomorphic sensor]] or, when employed in an array, the [[event camera]].<br />
<br />
The concept of neuromorphic systems can be extended to sensors (not just to computation). An example of this applied to detecting light is the retinomorphic sensor or, when employed in an array, the event camera.<br />
<br />
神经形态系统的概念可以扩展到传感器(而不仅仅是计算)。用于检测光线的一个例子是视网膜变形传感器，或者在阵列中使用的事件摄像机。<br />
<br />
==Ethical considerations ==<br />
==伦理问题==<br />
While the interdisciplinary concept of neuromorphic engineering is relatively new, many of the same ethical considerations apply to neuromorphic systems as apply to [[human-like machines]] and [[artificial intelligence]] in general. However, the fact that neuromorphic systems are designed to mimic a [[human brain]] gives rise to unique ethical questions surrounding their usage.<br />
<br />
While the interdisciplinary concept of neuromorphic engineering is relatively new, many of the same ethical considerations apply to neuromorphic systems as apply to human-like machines and artificial intelligence in general. However, the fact that neuromorphic systems are designed to mimic a human brain gives rise to unique ethical questions surrounding their usage.<br />
<br />
虽然神经形态工程的跨学科概念相对较新，但许多同样的伦理考虑适用于神经形态系统，就像适用于类人机器和一般人工智能一样。然而，神经形态系统是为了模仿人类大脑而设计的这一事实引起了围绕其使用的独特的伦理问题。<br />
<br />
However, the practical debate is that neuromorphic hardware as well as artificial "neural networks" are immensely simplified models of how the brain operates or processes information at a much lower [[complex system|complexity in terms of size and functional technology]] and a much more regular structure in terms of [[brain connectivity|connectivity]]. Comparing [[neuromorphic chip]]s to the brain is a very crude comparison similar to comparing a plane to a bird just because they both have wings and a tail. The fact is that neural cognitive systems are many orders of magnitude more [[Energy efficiency (physics)|energy-]] and compute-efficient than current state-of-the-art AI and neuromorphic engineering is an attempt to narrow this gap by inspiring from the brain's mechanism just like many engineering designs have [[bioengineering|bio-inspired features]].<br />
<br />
However, the practical debate is that neuromorphic hardware as well as artificial "neural networks" are immensely simplified models of how the brain operates or processes information at a much lower complexity in terms of size and functional technology and a much more regular structure in terms of connectivity. Comparing neuromorphic chips to the brain is a very crude comparison similar to comparing a plane to a bird just because they both have wings and a tail. The fact is that neural cognitive systems are many orders of magnitude more energy- and compute-efficient than current state-of-the-art AI and neuromorphic engineering is an attempt to narrow this gap by inspiring from the brain's mechanism just like many engineering designs have bio-inspired features.<br />
<br />
然而，实际上的争论是，神经形态硬件和人工“神经网络”是大脑如何运作或处理信息的极其简化的模型，在大小和功能技术方面的复杂性要低得多，在连接方面的结构则更加规则。将神经形态芯片与大脑进行比较是一种非常粗糙的比较，类似于仅仅因为一架飞机有翅膀和一条尾巴就将它与一只鸟进行比较。事实上，神经认知系统比当前最先进的人工智能具有更多的能量和计算效率，而神经形态工程是一种通过从大脑机制中激发灵感来缩小这种差距的尝试，就像许多工程设计具有生物启发的特征一样。数量级。<br />
<br />
===Democratic concerns===<br />
===公众担忧===<br />
Significant ethical limitations may be placed on neuromorphic engineering due to public perception.<ref>{{Cite report|url=https://ai100.stanford.edu/sites/g/files/sbiybj9861/f/ai_100_report_0831fnl.pdf|title=Artificial Intelligence and Life in 2030|author=2015 Study Panel|date=September 2016|work=One Hundred Year Study on Artificial Intelligence (AI100)|publisher=Stanford University}}</ref> Special [[Eurobarometer]] 382: Public Attitudes Towards Robots, a survey conducted by the European Commission, found that 60% of [[European Union]] citizens wanted a ban of robots in the care of children, the elderly, or the disabled. Furthermore, 34% were in favor of a ban on robots in education, 27% in healthcare, and 20% in leisure. The European Commission classifies these areas as notably “human.” The report cites increased public concern with robots that are able to mimic or replicate human functions. Neuromorphic engineering, by definition, is designed to replicate the function of the human brain.<ref name=":1">{{Cite web|url=http://ec.europa.eu/commfrontoffice/publicopinion/archives/ebs/ebs_382_en.pdf|title=Special Eurobarometer 382: Public Attitudes Towards Robots|last=European Commission|date=September 2012|website=European Commission}}</ref><br />
<br />
Significant ethical limitations may be placed on neuromorphic engineering due to public perception. Special Eurobarometer 382: Public Attitudes Towards Robots, a survey conducted by the European Commission, found that 60% of European Union citizens wanted a ban of robots in the care of children, the elderly, or the disabled. Furthermore, 34% were in favor of a ban on robots in education, 27% in healthcare, and 20% in leisure. The European Commission classifies these areas as notably “human.” The report cites increased public concern with robots that are able to mimic or replicate human functions. Neuromorphic engineering, by definition, is designed to replicate the function of the human brain.<br />
<br />
由于公众的认知，神经形态工程学可能会受到严重的伦理限制。欧盟委员会进行的一项调查发现，60% 的欧盟公民希望禁止照顾儿童、老人或残疾人的机器人。此外，34% 的人支持禁止机器人用于教育，27% 的人支持医疗保健，20% 的人支持休闲。欧盟委员会将这些地区明显归类为“人类”报告指出，公众越来越关注能够模仿或复制人类功能的机器人。神经形态工程，顾名思义，是为了复制人脑的功能而设计的。<br />
<br />
The democratic concerns surrounding neuromorphic engineering are likely to become even more profound in the future. The European Commission found that EU citizens between the ages of 15 and 24 are more likely to think of robots as human-like (as opposed to instrument-like) than EU citizens over the age of 55. When presented an image of a robot that had been defined as human-like, 75% of EU citizens aged 15–24 said it corresponded with the idea they had of robots while only 57% of EU citizens over the age of 55 responded the same way. The human-like nature of neuromorphic systems, therefore, could place them in the categories of robots many EU citizens would like to see banned in the future.<ref name=":1" /><br />
<br />
The democratic concerns surrounding neuromorphic engineering are likely to become even more profound in the future. The European Commission found that EU citizens between the ages of 15 and 24 are more likely to think of robots as human-like (as opposed to instrument-like) than EU citizens over the age of 55. When presented an image of a robot that had been defined as human-like, 75% of EU citizens aged 15–24 said it corresponded with the idea they had of robots while only 57% of EU citizens over the age of 55 responded the same way. The human-like nature of neuromorphic systems, therefore, could place them in the categories of robots many EU citizens would like to see banned in the future.<br />
<br />
围绕神经形态工程的民主关注可能在未来变得更加深刻。欧盟委员会(European Commission)发现，15至24岁的欧盟公民比55岁以上的欧盟公民更有可能认为机器人像人(而不是像仪器)。当看到一张被定义为“类人”的机器人图片时，年龄在15岁至24岁之间的欧盟公民中有75% 的人表示，这与他们对机器人的想法相符，而55岁以上的欧盟公民中只有57% 的人有同样的反应。因此，类似人类的神经形态系统，可以把它们归入许多欧盟公民希望在未来禁止使用的机器人类别。<br />
<br />
===Personhood===<br />
===人格问题===<br />
As neuromorphic systems have become increasingly advanced, some scholars{{who|date=August 2021}} have advocated for granting [[personhood]] rights to these systems. If the brain is what grants humans their personhood, to what extent does a neuromorphic system have to mimic the human brain to be granted personhood rights? Critics of technology development in the [[Human Brain Project]], which aims to advance brain-inspired computing, have argued that advancement in neuromorphic computing could lead to machine consciousness or personhood.<ref>{{Cite journal|last=Aicardi|first=Christine|date=September 2018|title=Accompanying technology development in the Human Brain Project: From foresight to ethics management|journal=Futures|volume=102|pages=114–124|doi=10.1016/j.futures.2018.01.005|doi-access=free}}</ref> If these systems are to be treated as people, critics argue, then many tasks humans perform using neuromorphic systems, including the act of termination of neuromorphic systems, may be morally impermissible as these acts would violate the autonomy of the neuromorphic systems.<ref>{{Cite journal|last=Lim|first=Daniel|date=2014-06-01|title=Brain simulation and personhood: a concern with the Human Brain Project|journal=Ethics and Information Technology|language=en|volume=16|issue=2|pages=77–89|doi=10.1007/s10676-013-9330-5|s2cid=17415814|issn=1572-8439}}</ref><br />
<br />
As neuromorphic systems have become increasingly advanced, some scholars have advocated for granting personhood rights to these systems. If the brain is what grants humans their personhood, to what extent does a neuromorphic system have to mimic the human brain to be granted personhood rights? Critics of technology development in the Human Brain Project, which aims to advance brain-inspired computing, have argued that advancement in neuromorphic computing could lead to machine consciousness or personhood. If these systems are to be treated as people, critics argue, then many tasks humans perform using neuromorphic systems, including the act of termination of neuromorphic systems, may be morally impermissible as these acts would violate the autonomy of the neuromorphic systems.<br />
<br />
随着神经形态系统的日益发展，一些学者主张赋予这些系统人格权。如果是大脑赋予了人类人格，那么神经形态系统在多大程度上必须模仿人类大脑才能被赋予人格权利？“人脑计划”旨在推进以大脑为灵感的计算机技术发展，该计划的批评者认为，神经形态计算机技术的进步可能导致机器意识或人格的形成。批评家认为，如果这些系统被当作人来对待，那么人类使用神经形态系统执行的许多任务，包括终止神经形态系统的行为，可能在道德上是不允许的，因为这些行为将违反神经形态系统的自主性。<br />
<br />
==Dual use (military applications)==<br />
==军民两用技术==<br />
The [[Joint Artificial Intelligence Center]], a branch of the U.S. military, is a center dedicated to the procurement and implementation of AI software and neuromorphic hardware for combat use. Specific applications include smart headsets/goggles and robots. JAIC intends to rely heavily on neuromorphic technology to connect "every fighter every shooter" within a network of neuromorphic-enabled units.<br />
<br />
The Joint Artificial Intelligence Center, a branch of the U.S. military, is a center dedicated to the procurement and implementation of AI software and neuromorphic hardware for combat use. Specific applications include smart headsets/goggles and robots. JAIC intends to rely heavily on neuromorphic technology to connect "every fighter every shooter" within a network of neuromorphic-enabled units.<br />
<br />
军民两用联合人工智能中心是美国军队的一个分支，专门从事采购和实施用于战斗的人工智能软件和神经形态硬件。具体应用包括智能耳机/护目镜和机器人。JAIC 打算严重依赖神经形态技术来连接神经形态单位网络中的“每个战士每个射手”。<br />
<br />
==Legal considerations==<br />
== 法律问题==<br />
Skeptics have argued that there is no way to apply the electronic personhood, the concept of personhood that would apply to neuromorphic technology, legally. In a letter signed by 285 experts in law, robotics, medicine, and ethics opposing a European Commission proposal to recognize “smart robots” as legal persons, the authors write, “A legal status for a robot can’t derive from the [[Natural person|Natural Person]] model, since the robot would then hold [[human rights]], such as the right to dignity, the right to its integrity, the right to remuneration or the right to citizenship, thus directly confronting the Human rights. This would be in contradiction with the [[Charter of Fundamental Rights of the European Union]] and the [[Convention for the Protection of Human Rights and Fundamental Freedoms]].”<ref>{{Cite web|url=http://www.robotics-openletter.eu/|title=Robotics Openletter {{!}} Open letter to the European Commission|language=fr-FR|access-date=2019-05-10}}</ref><br />
<br />
Skeptics have argued that there is no way to apply the electronic personhood, the concept of personhood that would apply to neuromorphic technology, legally. In a letter signed by 285 experts in law, robotics, medicine, and ethics opposing a European Commission proposal to recognize “smart robots” as legal persons, the authors write, “A legal status for a robot can’t derive from the Natural Person model, since the robot would then hold human rights, such as the right to dignity, the right to its integrity, the right to remuneration or the right to citizenship, thus directly confronting the Human rights. This would be in contradiction with the Charter of Fundamental Rights of the European Union and the Convention for the Protection of Human Rights and Fundamental Freedoms.”<br />
<br />
怀疑论者认为，在法律上没有办法应用电子人格，这个人格概念将适用于神经形态技术。在一封由285名法律、机器人、医学和伦理学专家签名的信中，作者们反对欧盟委员会承认“智能机器人”为法人的提议，他们写道，“机器人的法律地位不能从自然人模型中推导出来，因为机器人将拥有人权，如尊严权、完整权、报酬权或公民权，从而直接面对人权。这将有悖于《欧洲联盟基本权利宪章和《保护人权和基本自由公约》。”<br />
<br />
===Ownership and property rights===<br />
===所有权及财产权问题===<br />
There is significant legal debate around property rights and artificial intelligence. In ''Acohs Pty Ltd v. Ucorp Pty Ltd'', Justice Christopher Jessup of the [[Federal Court of Australia]] found that the [[source code]] for [[Material safety data sheets|Material Safety Data Sheets]] could not be [[Copyright law of Australia|copyrighted]] as it was generated by a [[software interface]] rather than a human author.<ref>{{Cite web|url=http://www.lavan.com.au/advice/intellectual_property_technology/copyright_in_source_code_and_digital_products|title=Copyright in source code and digital products|last=Lavan|website=Lavan|language=en|access-date=2019-05-10}}</ref> The same question may apply to neuromorphic systems: if a neuromorphic system successfully mimics a human brain and produces a piece of original work, who, if anyone, should be able to claim ownership of the work?<ref>{{cite journal |last1=Eshraghian|first1=Jason K. |title=Human Ownership of Artificial Creativity |journal=Nature Machine Intelligence |date=9 March 2020 |volume=2 |pages=157–160 |doi=10.1038/s42256-020-0161-x}}</ref><br />
<br />
There is significant legal debate around property rights and artificial intelligence. In Acohs Pty Ltd v. Ucorp Pty Ltd, Justice Christopher Jessup of the Federal Court of Australia found that the source code for Material Safety Data Sheets could not be copyrighted as it was generated by a software interface rather than a human author. The same question may apply to neuromorphic systems: if a neuromorphic system successfully mimics a human brain and produces a piece of original work, who, if anyone, should be able to claim ownership of the work?<br />
<br />
围绕财产权和人工智能有着重大的法律争论。在 Acohs Pty Ltd 诉 Ucorp Pty Ltd 一案中，澳大利亚联邦法院的克里斯托弗 · 杰瑟普法官发现，材料安全数据表的源代码不能受版权保护，因为它是由软件界面而不是人工作者生成的。同样的问题可能也适用于神经形态系统: 如果一个神经形态系统成功地模仿了人类的大脑并产生了一部原创作品，那么谁，如果有人，应该声称拥有这部作品的所有权？<br />
<br />
==Neuromemristive systems==<br />
==神经记忆电阻系统==<br />
Neuromemristive systems are a subclass of neuromorphic computing systems that focus on the use of [[memristors]] to implement [[neuroplasticity]]. While neuromorphic engineering focuses on mimicking biological behavior, neuromemristive systems focus on abstraction.<ref>{{Cite web|url=https://digitalops.sandia.gov/Mediasite/Play/a10cf6ceb55d47608bb8326dd00e46611d|title=002.08 N.I.C.E. Workshop 2014: Towards Intelligent Computing with Neuromemristive Circuits and Systems - Feb. 2014|website=digitalops.sandia.gov|access-date=2019-08-26}}</ref> For example, a neuromemristive system may replace the details of a [[Cerebral cortex|cortical]] microcircuit's behavior with an abstract neural network model.<ref>C. Merkel and D. Kudithipudi, "Neuromemristive extreme learning machines for pattern classification," ISVLSI, 2014.</ref><br />
<br />
Neuromemristive systems are a subclass of neuromorphic computing systems that focus on the use of memristors to implement neuroplasticity. While neuromorphic engineering focuses on mimicking biological behavior, neuromemristive systems focus on abstraction. For example, a neuromemristive system may replace the details of a cortical microcircuit's behavior with an abstract neural network model.C. Merkel and D. Kudithipudi, "Neuromemristive extreme learning machines for pattern classification," ISVLSI, 2014.<br />
<br />
神经记忆电阻系统是神经形态计算系统的一个亚类，主要研究利用记忆电阻器实现神经可塑性。神经形态工程的重点是模拟生物行为，而神经记忆电阻系统的重点是提取。例如，一个神经记忆系统可以用一个抽象的神经网络模型取代皮层微电路的行为细节。默克尔和 d. Kudithipudi，“用于模式分类的神经记忆极端学习机器”，ISVLSI，2014。<br />
<br />
There exist several neuron inspired threshold logic functions<ref name="Maan 1–13" /> implemented with memristors that have applications in high level [[pattern recognition]] applications. Some of the applications reported recently include [[speech recognition]],<ref>{{Cite journal|title = Memristor pattern recogniser: isolated speech word recognition|journal = Electronics Letters|pages = 1370–1372|volume = 51|issue = 17|doi = 10.1049/el.2015.1428|first1 = A.K.|last1 = Maan|first2 = A.P.|last2 = James|first3 = S.|last3 = Dimitrijev|year = 2015|bibcode = 2015ElL....51.1370M|hdl = 10072/140989|s2cid = 61454815|url = https://semanticscholar.org/paper/48d3ab11ec6e213b62f11eedcfb7b7febb058674|hdl-access = free}}</ref> [[face recognition]]<ref>{{Cite journal|title = Memristive Threshold Logic Face Recognition|journal = Procedia Computer Science|date = 2014-01-01|pages = 98–103|volume = 41|series = 5th Annual International Conference on Biologically Inspired Cognitive Architectures, 2014 BICA|doi = 10.1016/j.procs.2014.11.090|first1 = Akshay Kumar|last1 = Maan|first2 = Dinesh S.|last2 = Kumar|first3 = Alex Pappachen|last3 = James|doi-access = free}}</ref> and [[object recognition]].<ref>{{Cite journal|title = Memristive Threshold Logic Circuit Design of Fast Moving Object Detection|journal = IEEE Transactions on Very Large Scale Integration (VLSI) Systems|date = 2015-10-01|issn = 1063-8210|pages = 2337–2341|volume = 23|issue = 10|doi = 10.1109/TVLSI.2014.2359801|first1 = A.K.|last1 = Maan|first2 = D.S.|last2 = Kumar|first3 = S.|last3 = Sugathan|first4 = A.P.|last4 = James|arxiv = 1410.1267|s2cid = 9647290}}</ref> They also find applications in replacing conventional digital logic gates.<ref>{{Cite journal|title = Resistive Threshold Logic|journal = IEEE Transactions on Very Large Scale Integration (VLSI) Systems|date = 2014-01-01|issn = 1063-8210|pages = 190–195|volume = 22|issue = 1|doi = 10.1109/TVLSI.2012.2232946|first1 = A.P.|last1 = James|first2 = L.R.V.J.|last2 = Francis|first3 = D.S.|last3 = Kumar|arxiv = 1308.0090|s2cid = 7357110}}</ref><ref>{{Cite journal|title = Threshold Logic Computing: Memristive-CMOS Circuits for Fast Fourier Transform and Vedic Multiplication|journal = IEEE Transactions on Very Large Scale Integration (VLSI) Systems|date = 2015-11-01|issn = 1063-8210|pages = 2690–2694|volume = 23|issue = 11|doi = 10.1109/TVLSI.2014.2371857|first1 = A.P.|last1 = James|first2 = D.S.|last2 = Kumar|first3 = A.|last3 = Ajayan|arxiv = 1411.5255|s2cid = 6076956}}</ref><br />
<br />
There exist several neuron inspired threshold logic functions implemented with memristors that have applications in high level pattern recognition applications. Some of the applications reported recently include speech recognition, face recognition and object recognition. They also find applications in replacing conventional digital logic gates.<br />
<br />
利用记忆电阻器实现的神经元启发阈值逻辑函数在高级模式识别中有着广泛的应用。最近报道的一些应用包括语音识别、人脸识别和物体识别。它们还可以用来取代传统的数字逻辑门。<br />
<br />
For ideal passive memristive circuits there is an exact equation (Caravelli-Traversa-[[Di Ventra]] equation) for the internal memory of the circuit:<ref>{{cite journal |last=Caravelli |display-authors=etal|arxiv=1608.08651 |title=The complex dynamics of memristive circuits: analytical results and universal slow relaxation |year=2017 |doi=10.1103/PhysRevE.95.022140 |pmid= 28297937 |volume=95 |issue= 2 |pages= 022140 |journal=Physical Review E|bibcode=2017PhRvE..95b2140C |s2cid=6758362}}</ref><br />
<br />
For ideal passive memristive circuits there is an exact equation (Caravelli-Traversa-Di Ventra equation) for the internal memory of the circuit:<br />
<br />
对于理想的无源记忆电路，电路的内部记忆有一个精确的方程(Caravelli-Traversa-Di Ventra 方程) :<br />
<br />
:<math> \frac{d}{dt} \vec{W} = \alpha \vec{W}-\frac{1}{\beta} (I+\xi \Omega W)^{-1} \Omega \vec S </math><br />
<br />
:\frac{d}{dt} \vec{W} = \alpha \vec{W}-\frac{1}{\beta} (I+\xi \Omega W)^{-1} \Omega \vec S<br />
<br />
:\frac{d}{dt} \vec{W} = \alpha \vec{W}-\frac{1}{\beta} (I+\xi \Omega W)^{-1} \Omega \vec S<br />
<br />
as a function of the properties of the physical memristive network and the external sources. In the equation above, <math>\alpha</math> is the "forgetting" time scale constant, <math> \xi=r-1</math> and <math>r=\frac{R_\text{off}}{R_\text{on}}</math> is the ratio of ''off'' and ''on'' values of the limit resistances of the memristors, <math> \vec S </math> is the vector of the sources of the circuit and <math>\Omega</math> is a projector on the fundamental loops of the circuit. The constant <math>\beta</math> has the dimension of a voltage and is associated to the properties of the [[memristor]]; its physical origin is the charge mobility in the conductor. The diagonal matrix and vector <math>W=\operatorname{diag}(\vec W)</math> and <math>\vec W</math> respectively, are instead the internal value of the memristors, with values between 0 and 1. This equation thus requires adding extra constraints on the memory values in order to be reliable.<br />
<br />
<nowiki>as a function of the properties of the physical memristive network and the external sources. In the equation above, \alpha is the "forgetting" time scale constant, \xi=r-1 and r=\frac{R_\text{off}}{R_\text{on}} is the ratio of off and on values of the limit resistances of the memristors, \vec S is the vector of the sources of the circuit and \Omega is a projector on the fundamental loops of the circuit. The constant \beta has the dimension of a voltage and is associated to the properties of the memristor; its physical origin is the charge mobility in the conductor. The diagonal matrix and vector W=\operatorname{diag}(\vec W) and \vec W respectively, are instead the internal value of the memristors, with values between 0 and 1. This equation thus requires adding extra constraints on the memory values in order to be reliable.</nowiki><br />
<br />
<nowiki>作为物理记忆网络和外部源的性质的函数。在上述方程中，α 是“遗忘”时间尺度常数，xi = r-1，r = frac { r _ text { off }{ on }{ r _ text { on }}是记忆电阻器极限电阻的开关和开关值之比，vec s 是电路源的矢量，Omega 是电路基本环路的投影仪。常数 β 具有电压的尺寸，与记忆电阻器的特性有关; 它的物理起源是导体中的电荷迁移率。对角矩阵和向量 w = 操作者名{ diag }(vec w)和 vec w 分别是记忆电阻器的内值，值在0到1之间。因此，这个等式需要在内存值上添加额外的约束，以保证可靠性。</nowiki><br />
<br />
==See also==<br />
==相关词条==<br />
{{Columns-list|colwidth=18em|<br />
* [[AI accelerator (computer hardware)|AI accelerator]]<br />
* [[Artificial brain]]<br />
* [[Biomorphic]]<br />
* [[Cognitive computer]]<br />
* [[Computation and Neural Systems]]<br />
* [[Differentiable programming]]<br />
* [[Event camera]]<br />
* [[Neurorobotics]]<br />
* [[Optical flow sensor]]<br />
* [[Physical neural network]]<br />
* [[SpiNNaker]]<br />
* [[SyNAPSE]]<br />
* [[Retinomorphic sensor]]<br />
* [[Vision chip]]<br />
* [[Vision processing unit]]<br />
* [[Zeroth (software)]]<br />
* [[Hardware for artificial intelligence]]<br />
}}<br />
<br />
{{Portal bar|Electronics}}<br />
<br />
==References==<br />
{{Reflist|40em}}<br />
<br />
==External links==<br />
<!--======================== {{No more links}} ============================<br />
| PLEASE BE CAUTIOUS IN ADDING MORE LINKS TO THIS ARTICLE. Wikipedia |<br />
| is not a collection of links nor should it be used for advertising. |<br />
| |<br />
| Excessive or inappropriate links WILL BE DELETED. |<br />
| See [[Wikipedia:External links]] & [[Wikipedia:Spam]] for details. |<br />
| |<br />
| If there are already plentiful links, please propose additions or |<br />
| replacements on this article's discussion page, or submit your link |<br />
| to the relevant category at the Open Directory Project (dmoz.org) |<br />
| and link back to that category using the {{dmoz}} template. |<br />
======================={{No more links}}=============================--><br />
{{Commons category|Neuromorphic Engineering}}<br />
*[https://web.archive.org/web/20150727034331/http://ine-web.org/workshops/workshops-overview Telluride Neuromorphic Engineering Workshop]<br />
*[https://archive.today/20130115190057/http://capocaccia.ethz.ch/ CapoCaccia Cognitive Neuromorphic Engineering Workshop]<br />
*[https://web.archive.org/web/20190716132350/http://www.ine-web.org/ Institute of Neuromorphic Engineering]<br />
*[http://www.ine-news.org/ INE news site].<br />
*[http://www.frontiersin.org/neuromorphic_engineering Frontiers in Neuromorphic Engineering Journal]<br />
*[http://www.cns.caltech.edu/ Computation and Neural Systems] department at the [[California Institute of Technology]].<br />
*[http://www.humanbrainproject.eu/ Human Brain Project official site]<br />
*[https://www.the-scientist.com/features/building-a-silicon-brain-65738 Building a Silicon Brain:] Computer chips based on biological neurons may help simulate larger and more-complex brain models. May 1, 2019. SANDEEP RAVINDRAN<br />
<br />
<br />
*Telluride Neuromorphic Engineering Workshop<br />
*CapoCaccia Cognitive Neuromorphic Engineering Workshop<br />
*Institute of Neuromorphic Engineering<br />
*INE news site.<br />
*Frontiers in Neuromorphic Engineering Journal<br />
*Computation and Neural Systems department at the California Institute of Technology.<br />
*Human Brain Project official site<br />
*Building a Silicon Brain: Computer chips based on biological neurons may help simulate larger and more-complex brain models. May 1, 2019. SANDEEP RAVINDRAN<br />
<br />
=<nowiki>外部链接</nowiki>= <br />
*碲化物神经形态工程工作室<br />
*CapoCaccia 认知神经形态工程工作室<br />
*神经形态工程研究所<br />
*INE 新闻站点。<br />
*《神经形态工程学前沿》<br />
*加州理工学院计算与神经系统系。<br />
*人脑项目官方网站<br />
*打造硅脑: 基于生物神经元的计算机芯片可能有助于模拟更大、更复杂的大脑模型。2019年5月1日。SANDEEP RAVINDRAN<br />
<br />
{{Differentiable computing}}<br />
{{Authority control}}<br />
<br />
[[Category:Electrical engineering]]<br />
[[Category:Neuroscience]]<br />
[[分类:AI accelerators|*]]<br />
[[Category:Artificial intelligence]]<br />
[[Category:Robotics]]<br />
<br />
Category:Electrical engineering<br />
Category:Neuroscience<br />
*<br />
Category:Artificial intelligence<br />
Category:Robotics<br />
<br />
类别: 电气工程类别: 神经科学 * 类别: 人工智能类别: 机器人<br />
<br />
<noinclude><br />
<br />
<small>This page was moved from [[wikipedia:en:Neuromorphic engineering]]. Its edit history can be viewed at [[神经形态计算/edithistory]]</small></noinclude><br />
<br />
[[Category:待整理页面]]<br />
<br />
<references /></div>Shenky20https://wiki.swarma.org/index.php?title=%E7%A5%9E%E7%BB%8F%E5%BD%A2%E6%80%81%E8%AE%A1%E7%AE%97&diff=30175神经形态计算2022-04-13T07:53:36Z<p>Shenky20：</p>
<hr />
<div><br />
此词条由神经动力学模型读书会词条梳理志愿者Shenky20翻译审校，未经专家审校，带来阅读不便，请见谅。<br />
<br />
{{Use American English|date = January 2019}}<br />
{{Use mdy dates|date = January 2019}}<br />
<br />
'''Neuromorphic engineering''', also known as '''neuromorphic computing''',<ref name=":3">{{Cite journal | doi = 10.1145/2601069| title = Neuromorphic computing gets ready for the (really) big time| journal = [[Communications of the ACM]]| volume = 57| issue = 6| year = 2014| last1 = Monroe | first1 = D. | pages = 13–15| s2cid = 20051102}}</ref><ref name=":4">{{Cite journal | doi = 10.1088/0957-4484/21/17/175202| pmid = 20368686| title = Nanotube devices based crossbar architecture: Toward neuromorphic computing| journal = Nanotechnology| volume = 21| issue = 17| pages = 175202| year = 2010| last1 = Zhao | first1 = W. S. | last2 = Agnus | first2 = G. | last3 = Derycke | first3 = V. | last4 = Filoramo | first4 = A. | last5 = Bourgoin | first5 = J. -P. | last6 = Gamrat | first6 = C. | bibcode = 2010Nanot..21q5202Z| url = https://zenodo.org/record/3428659}}</ref><ref name="humanbrainproject">{{YouTube|id=6RoiZ90mGfw|title=The Human Brain Project SP 9: Neuromorphic Computing Platform}}</ref> is the use of [[very-large-scale integration]] (VLSI) systems containing electronic [[analog circuit]]s to mimic neuro-biological architectures present in the nervous system. A neuromorphic computer/chip is any device that uses physical artificial neurons (made from silicon) to do computations.<ref name=":5">{{cite journal|last1=Mead|first1=Carver|title=Neuromorphic electronic systems|journal=Proceedings of the IEEE|date=1990|volume=78|issue=10|pages=1629–1636|doi=10.1109/5.58356|url=https://authors.library.caltech.edu/53090/1/00058356.pdf}}</ref><ref name=":2" /> In recent times, the term ''neuromorphic'' has been used to describe analog, digital, mixed-mode analog/digital VLSI, and software systems that implement models of [[neural system]]s (for [[perception]], [[motor control]], or [[multisensory integration]]). The implementation of neuromorphic computing on the hardware level can be realized by oxide-based [[memristor]]s,<ref name="Maan 1–13">{{Cite journal|last1=Maan|first1=A. K.|last2=Jayadevi|first2=D. A.|last3=James|first3=A. P.|date=2016-01-01|title=A Survey of Memristive Threshold Logic Circuits|journal=IEEE Transactions on Neural Networks and Learning Systems|volume=PP|issue=99|pages=1734–1746|doi=10.1109/TNNLS.2016.2547842|pmid=27164608|issn=2162-237X|arxiv=1604.07121|bibcode=2016arXiv160407121M|s2cid=1798273}}</ref> spintronic memories, threshold switches, and [[transistor]]s.<ref name=":6">{{Cite journal|title = Mott Memory and Neuromorphic Devices|journal = Proceedings of the IEEE|date = 2015-08-01|issn = 0018-9219|pages = 1289–1310|volume = 103|issue = 8|doi = 10.1109/JPROC.2015.2431914|first1 = You|last1 = Zhou|first2 = S.|last2 = Ramanathan|s2cid = 11347598|url = https://zenodo.org/record/895565}}</ref><ref name=":2">{{Cite document|title=Neuromorphic Circuits With Neural Modulation Enhancing the Information Content of Neural Signaling {{!}} International Conference on Neuromorphic Systems 2020|language=EN|doi=10.1145/3407197.3407204|s2cid=220794387}}</ref> Training software-based neuromorphic systems of spiking neural networks can be achieved using error backpropagation, e.g., using [[Python (programming language)|Python]] based frameworks such as snnTorch,<ref name=":7">{{cite journal |last1=Eshraghian|first1=Jason K.|last2=Ward|first2=Max|last3=Neftci |first3=Emre|last4=Wang|first4=Xinxin|last5=Lenz|first5=Gregor|last6=Dwivedi|first6=Girish|last7=Bennamoun|first7=Mohammed|last8=Jeong|first8=Doo Seok|last9=Lu|first9=Wei D.|title=Training Spiking Neural Networks Using Lessons from Deep Learning |date=1 October 2021 |arxiv=2109.12894 }}</ref> or using canonical learning rules from the biological learning literature, e.g., using BindsNet.<ref name=":8">{{Cite web | url=https://github.com/Hananel-Hazan/bindsnet | title=Hananel-Hazan/bindsnet: Simulation of spiking neural networks (SNNs) using PyTorch.| date=31 March 2020}}</ref><br />
<br />
<br />
'''<font color="#ff8000">神经形态工程Neuromorphic engineering</font>'''(也称为'''<font color="#ff8000">神经形态计算Neuromorphic computing</font>''')<ref name=":3" /><ref name=":4" /><ref name="humanbrainproject" />是指使用包含电子'''<font color="#ff8000">模拟电路Analog circuit</font>'''的'''<font color="#ff8000">超大规模集成电路Very-large-scale integration</font>'''系统来模拟神经系统中生理结构的研究方法。神经形态计算机或神经形态芯片包括任何使用由硅制成的人造神经元进行计算的设备。<ref name=":5" /><ref name=":2" />近年来，神经形态学（neuromorphic）这个术语被用来描述能够实现'''<font color="#ff8000">神经系统Neural system</font>'''模型功能(如'''<font color="#ff8000">感知Perception</font>'''、'''<font color="#ff8000">运动控制Motor control</font>'''，'''<font color="#ff8000">多感官整合Multisensory integration</font>'''等)的模拟、数字、模拟/数字混合模式超大规模集成电路和软件系统。神经形态计算的硬件实现可以通过基于氧化物的'''<font color="#ff8000">记忆电阻器Memristor</font>'''、<ref name="Maan 1–13" />自旋电子存储器、阈值开关和'''<font color="#ff8000">晶体管Transistor</font>'''来实现。<ref name=":6" /><ref name=":2" />对基于软件的脉冲神经网络系统的训练可以通过误差反向传播机制来实现，例如，使用snnTorch等基于Python的框架，<ref name=":7" />或使用BindsNet等典型的受生物启发的学习模式。<ref name=":8" /><br />
<br />
A key aspect of neuromorphic engineering is understanding how the morphology of individual neurons, circuits, applications, and overall architectures creates desirable computations, affects how information is represented, influences robustness to damage, incorporates learning and development, adapts to local change (plasticity), and facilitates evolutionary change.<br />
<br />
<br />
神经形态工程领域的一个关键问题，就是理解'''<font color="#32CD32">单个神经元形态、神经回路、应用和整体结构</font>'''如何产生理想的计算，如何影响信息的表示和对破坏的鲁棒性，如何整合学习和发展，如何适应局部变化(可塑性) 并促进逐渐发展的变化。<br />
<br />
Neuromorphic engineering is an interdisciplinary subject that takes inspiration from [[biology]], [[physics]], [[mathematics]], [[computer science]], and [[electronic engineering]]<ref name=":2" /> to design artificial neural systems, such as [[Machine vision|vision systems]], head-eye systems, auditory processors, and autonomous robots, whose physical architecture and design principles are based on those of biological nervous systems.<ref name=":9">{{Cite journal | doi = 10.1155/2012/705483| title = Qualitative Functional Decomposition Analysis of Evolved Neuromorphic Flight Controllers| journal = Applied Computational Intelligence and Soft Computing| volume = 2012| pages = 1–21| year = 2012| last1 = Boddhu | first1 = S. K. | last2 = Gallagher | first2 = J. C. | doi-access = free}}</ref> It was developed by [[Carver Mead]]<ref name=":10">{{cite web|title=carver mead website|last1=Mead|first1=Carver|url=http://www.carvermead.caltech.edu/index.html|website=carvermead}}</ref> in the late 1980s.<br />
<br />
<br />
神经形态工程是以'''<font color="#ff8000">生物学Biology</font>'''、'''<font color="#ff8000">物理学Physics</font>'''、'''<font color="#ff8000">数学Mathematics</font>'''、'''<font color="#ff8000">计算机科学Computer science</font>'''和'''<font color="#ff8000">电子工程Electronic engineering<ref name=":2" /></font>'''等学科为基础，设计人工神经系统（如'''<font color="#ff8000">视觉系统Vision systems</font>'''、头眼系统、听觉处理器以及物理结构和设计原则都受启发于生物神经系统的自主机器人）的一门交叉学科。<ref name=":9" />20世纪80年代后期，卡弗·米德极大地推动了神经形态工程领域的发展。<ref name=":10" /><br />
<br />
==Neurological inspiration==<br />
==神经生物学启发==<br />
Neuromorphic engineering is set apart by the inspiration it takes from what we know about the structure and operations of the [[brain]]. Neuromorphic engineering translates what we know about the brain's function into computer systems. Work has mostly focused on replicating the analog nature of [[biological computation]] and the role of [[neuron]]s in [[cognition]].<br />
<br />
Neuromorphic engineering is set apart by the inspiration it takes from what we know about the structure and operations of the brain. Neuromorphic engineering translates what we know about the brain's function into computer systems. Work has mostly focused on replicating the analog nature of biological computation and the role of neurons in cognition.<br />
<br />
神经学灵感工程学是从我们对大脑结构和运作的了解中获得灵感而形成的。神经形态工程学将我们对大脑功能的了解转化为计算机系统。工作主要集中在复制生物计算的模拟特性和神经元在认知中的作用。<br />
<br />
The biological processes of neurons and their [[synapse]]s are dauntingly complex, and thus very difficult to artificially simulate. A key feature of biological brains is that all of the processing in neurons use analog chemical signals. This makes it hard to replicate brains in computers because the current generation of computers is completely digital. However, the characteristics of these parts can be abstracted into mathematical functions that closely capture the essence of the neuron's operations.<br />
<br />
The biological processes of neurons and their synapses are dauntingly complex, and thus very difficult to artificially simulate. A key feature of biological brains is that all of the processing in neurons use analog chemical signals. This makes it hard to replicate brains in computers because the current generation of computers is completely digital. However, the characteristics of these parts can be abstracted into mathematical functions that closely capture the essence of the neuron's operations.<br />
<br />
神经元及其突触的生物学过程极其复杂，难以人工模拟。生物大脑的一个关键特征是，神经元的所有处理过程都使用模拟化学信号。这使得很难在计算机中复制大脑，因为目前的计算机是完全数字化的。然而，这些部分的特征可以抽象为数学函数，这些函数紧密地抓住了神经元运行的本质。<br />
<br />
The goal of neuromorphic computing is not to perfectly mimic the brain and all of its functions, but instead to extract what is known of its structure and operations to be used in a practical computing system. No neuromorphic system will claim nor attempt to reproduce every element of neurons and synapses, but all adhere to the idea that computation is highly [[distributed processing|distributed]] throughout a series of small computing elements analogous to a neuron. While this sentiment is standard, researchers chase this goal with different methods.<ref>{{Cite journal | doi = 10.1088/1741-2560/13/5/051001| title = Large-scale neuromorphic computing systems| journal = Journal of Neural Engineering| volume = 13| pages = 1–15| year = 2016| last1 = Furber | first1 = Steve| issue = 5| pmid = 27529195| bibcode = 2016JNEng..13e1001F| doi-access = free}}</ref><br />
<br />
The goal of neuromorphic computing is not to perfectly mimic the brain and all of its functions, but instead to extract what is known of its structure and operations to be used in a practical computing system. No neuromorphic system will claim nor attempt to reproduce every element of neurons and synapses, but all adhere to the idea that computation is highly distributed throughout a series of small computing elements analogous to a neuron. While this sentiment is standard, researchers chase this goal with different methods.<br />
<br />
神经形态计算的目标不是完美地模拟大脑及其所有功能，而是提取已知的大脑结构和操作，用于实际的计算系统。没有哪个神经形态学系统会声称或试图复制神经元和突触的每一个元素，但所有人都坚持这样的观点，即计算是高度分布在一系列类似于神经元的小型计算元素中的。虽然这种情绪是标准的，但研究人员用不同的方法追求这一目标。<br />
<br />
==Examples==<br />
==案例==<br />
As early as 2006, researchers at [[Georgia Tech]] published a field programmable neural array.<ref>{{Cite book|title = A field programmable neural array|last1 = Farquhar|first1 = Ethan|date = May 2006|journal = IEEE International Symposium on Circuits and Systems|pages = 4114–4117|last2 = Hasler|first2 = Paul.|doi = 10.1109/ISCAS.2006.1693534|isbn = 978-0-7803-9389-9|s2cid = 206966013}}</ref> This chip was the first in a line of increasingly complex arrays of floating gate transistors that allowed programmability of charge on the gates of [[MOSFET]]s to model the channel-ion characteristics of neurons in the brain and was one of the first cases of a silicon programmable array of neurons.<br />
<br />
As early as 2006, researchers at Georgia Tech published a field programmable neural array. This chip was the first in a line of increasingly complex arrays of floating gate transistors that allowed programmability of charge on the gates of MOSFETs to model the channel-ion characteristics of neurons in the brain and was one of the first cases of a silicon programmable array of neurons.<br />
<br />
早在2006年，佐治亚理工学院的研究人员就发表了现场可编程神经阵列。这种芯片是一系列越来越复杂的浮栅晶体管阵列中的第一个，这些晶体管可以在 mosfet 的栅极上编程来模拟大脑中神经元的通道离子特性，这也是硅可编程神经元阵列的首批案例之一。<br />
<br />
In November 2011, a group of [[MIT]] researchers created a computer chip that mimics the analog, ion-based communication in a synapse between two neurons using 400 transistors and standard [[CMOS]] manufacturing techniques.<ref>{{cite web|title=MIT creates "brain chip"|url=http://www.extremetech.com/extreme/105067-mit-creates-brain-chip|access-date=4 December 2012}}</ref><ref name="Neuromorphic silicon paper">{{cite journal|title=Neuromorphic silicon neurons and large-scale neural networks: challenges and opportunities|doi=10.3389/fnins.2011.00108|pmid=21991244|pmc=3181466|volume=5|pages=108|journal=Frontiers in Neuroscience|year=2011|last1=Poon|first1=Chi-Sang|last2=Zhou|first2=Kuan|doi-access=free}}</ref><br />
<br />
In November 2011, a group of MIT researchers created a computer chip that mimics the analog, ion-based communication in a synapse between two neurons using 400 transistors and standard CMOS manufacturing techniques.<br />
<br />
2011年11月，麻省理工学院的一组研究人员发明了一种计算机芯片，该芯片使用400个晶体管和标准的 CMOS 制造技术，在两个神经元之间的突触中模拟模拟基于离子的通讯。<br />
<br />
In June 2012, [[spintronic]] researchers at [[Purdue University]] presented a paper on the design of a neuromorphic chip using [[Spin valve|lateral spin valve]]s and [[memristor]]s. They argue that the architecture works similarly to neurons and can therefore be used to test methods of reproducing the brain's processing. In addition, these chips are significantly more energy-efficient than conventional ones.<ref name="Spin Devices Prop">{{Cite arXiv|title=Proposal For Neuromorphic Hardware Using Spin Devices|eprint=1206.3227|last1=Sharad|first1=Mrigank|last2=Augustine|first2=Charles|last3=Panagopoulos|first3=Georgios|last4=Roy|first4=Kaushik|class=cond-mat.dis-nn|year=2012}}</ref><br />
<br />
In June 2012, spintronic researchers at Purdue University presented a paper on the design of a neuromorphic chip using lateral spin valves and memristors. They argue that the architecture works similarly to neurons and can therefore be used to test methods of reproducing the brain's processing. In addition, these chips are significantly more energy-efficient than conventional ones.<br />
<br />
2012年6月，普渡大学的自旋电子学研究人员发表了一篇关于利用侧向自旋阀和记忆电阻器设计神经形态芯片的论文。他们认为，这种结构的工作原理与神经元相似，因此可以用来测试复制大脑处理过程的方法。此外，这些芯片明显比传统芯片更节能。<br />
<br />
Research at [[HP Labs]] on Mott memristors has shown that while they can be non-[[Volatile memory|volatile]], the volatile behavior exhibited at temperatures significantly below the [[phase transition]] temperature can be exploited to fabricate a [[neuristor]],<ref name=":0" /> a biologically-inspired device that mimics behavior found in neurons.<ref name=":0">{{Cite journal | doi = 10.1038/nmat3510| pmid = 23241533| title = A scalable neuristor built with Mott memristors| journal = Nature Materials| volume = 12| issue = 2| pages = 114–7| year = 2012| last1 = Pickett | first1 = M. D. | last2 = Medeiros-Ribeiro | first2 = G. | last3 = Williams | first3 = R. S. | bibcode = 2013NatMa..12..114P| s2cid = 16271627| url = https://semanticscholar.org/paper/b6ba6f496ace2b947f111059663e76bb60e9efeb}}</ref> In September 2013, they presented models and simulations that show how the spiking behavior of these neuristors can be used to form the components required for a [[Turing machine]].<ref>{{cite journal|doi=10.1088/0957-4484/24/38/384002|title=Phase transitions enable computational universality in neuristor-based cellular automata|author1=Matthew D Pickett|author2=R Stanley Williams|name-list-style=amp|date=September 2013|publisher=IOP Publishing Ltd|journal=Nanotechnology|volume=24|issue=38|pmid=23999059|bibcode=2013Nanot..24L4002P|at=384002}}</ref><br />
<br />
Research at HP Labs on Mott memristors has shown that while they can be non-volatile, the volatile behavior exhibited at temperatures significantly below the phase transition temperature can be exploited to fabricate a neuristor, a biologically-inspired device that mimics behavior found in neurons. In September 2013, they presented models and simulations that show how the spiking behavior of these neuristors can be used to form the components required for a Turing machine.<br />
<br />
惠普实验室在莫特记忆电阻器上的研究表明，尽管它们可以是非挥发性的，但是在相变温度以下显示的挥发性行为可以被用来制造神经元电阻器，这是一种模仿神经元行为的生物设备。2013年9月，他们展示了模型和仿真，展示了这些神经元的尖峰行为是如何被用来形成图灵机所需的元件的。<br />
<br />
[[Neurogrid]], built by ''Brains in Silicon'' at [[Stanford University]],<ref>{{cite journal|last1=Boahen|first1=Kwabena|title=Neurogrid: A Mixed-Analog-Digital Multichip System for Large-Scale Neural Simulations|journal=Proceedings of the IEEE|date=24 April 2014|volume=102|issue=5|pages=699–716|doi=10.1109/JPROC.2014.2313565|s2cid=17176371}}</ref> is an example of hardware designed using neuromorphic engineering principles. The circuit board is composed of 16 custom-designed chips, referred to as NeuroCores. Each NeuroCore's analog circuitry is designed to emulate neural elements for 65536 neurons, maximizing energy efficiency. The emulated neurons are connected using digital circuitry designed to maximize spiking throughput.<ref>{{cite journal|doi=10.1038/503022a|pmid = 24201264|title = Neuroelectronics: Smart connections|journal = Nature|volume = 503|issue = 7474|pages = 22–4|year = 2013|last1 = Waldrop|first1 = M. Mitchell|bibcode = 2013Natur.503...22W|doi-access = free}}</ref><ref>{{cite journal|doi=10.1109/JPROC.2014.2313565|title = Neurogrid: A Mixed-Analog-Digital Multichip System for Large-Scale Neural Simulations|journal = Proceedings of the IEEE|volume = 102|issue = 5|pages = 699–716|year = 2014|last1 = Benjamin|first1 = Ben Varkey|last2 = Peiran Gao|last3 = McQuinn|first3 = Emmett|last4 = Choudhary|first4 = Swadesh|last5 = Chandrasekaran|first5 = Anand R.|last6 = Bussat|first6 = Jean-Marie|last7 = Alvarez-Icaza|first7 = Rodrigo|last8 = Arthur|first8 = John V.|last9 = Merolla|first9 = Paul A.|last10 = Boahen|first10 = Kwabena|s2cid = 17176371}}</ref><br />
<br />
Neurogrid, built by Brains in Silicon at Stanford University, is an example of hardware designed using neuromorphic engineering principles. The circuit board is composed of 16 custom-designed chips, referred to as NeuroCores. Each NeuroCore's analog circuitry is designed to emulate neural elements for 65536 neurons, maximizing energy efficiency. The emulated neurons are connected using digital circuitry designed to maximize spiking throughput.<br />
<br />
神经网格是由斯坦福大学硅谷大脑公司建立的，是一个使用神经形态工程原理设计硬件的例子。该电路板由16个定制设计的芯片组成，称为 NeuroCores。每个 NeuroCore 的模拟电路被设计为模拟65536个神经元的神经元元件，最大限度地提高能量效率。模拟的神经元通过设计的数字电路连接，以最大化脉冲吞吐量。<br />
<br />
A research project with implications for neuromorphic engineering is the [[Human Brain Project]] that is attempting to simulate a complete human brain in a supercomputer using biological data. It is made up of a group of researchers in neuroscience, medicine, and computing.<ref>{{cite web|title=Involved Organizations|url=http://www.humanbrainproject.eu/partners.html|access-date=22 February 2013|url-status=dead|archive-url=https://web.archive.org/web/20130302142627/http://www.humanbrainproject.eu/partners.html|archive-date=2 March 2013}}</ref> [[Henry Markram]], the project's co-director, has stated that the project proposes to establish a foundation to explore and understand the brain and its diseases, and to use that knowledge to build new computing technologies. The three primary goals of the project are to better understand how the pieces of the brain fit and work together, to understand how to objectively diagnose and treat brain diseases, and to use the understanding of the human brain to develop neuromorphic computers. That the simulation of a complete human brain will require a supercomputer a thousand times more powerful than today's encourages the current focus on neuromorphic computers.<ref>{{cite web|title=Human Brain Project|url=http://www.humanbrainproject.eu|access-date=22 February 2013}}</ref> $1.3 billion has been allocated to the project by The [[European Commission]].<ref>{{cite web|title=The Human Brain Project and Recruiting More Cyberwarriors|url=http://www.marketplace.org/topics/tech/human-brain-project-and-recruiting-more-cyberwarriors|access-date=22 February 2013|date=January 29, 2013}}</ref><br />
<br />
A research project with implications for neuromorphic engineering is the Human Brain Project that is attempting to simulate a complete human brain in a supercomputer using biological data. It is made up of a group of researchers in neuroscience, medicine, and computing. Henry Markram, the project's co-director, has stated that the project proposes to establish a foundation to explore and understand the brain and its diseases, and to use that knowledge to build new computing technologies. The three primary goals of the project are to better understand how the pieces of the brain fit and work together, to understand how to objectively diagnose and treat brain diseases, and to use the understanding of the human brain to develop neuromorphic computers. That the simulation of a complete human brain will require a supercomputer a thousand times more powerful than today's encourages the current focus on neuromorphic computers. $1.3 billion has been allocated to the project by The European Commission.<br />
<br />
一个对神经形态工程有影响的研究项目是人脑项目，它试图用生物数据在超级计算机中模拟完整的人脑。它由一群神经科学、医学和计算机科学的研究人员组成。该项目的联合主管亨利•马克拉姆(Henry Markram)表示，该项目提议建立一个基础，以探索和了解大脑及其疾病，并利用这些知识构建新的计算机技术。这个项目的三个主要目标是: 更好地理解大脑的各个部分是如何相互匹配和协同工作的; 理解如何客观地诊断和治疗脑部疾病; 以及利用对人类大脑的理解来开发神经形态计算机。模拟一个完整的人类大脑需要一台比现在强大一千倍的超级计算机，这鼓励了当前对神经形态计算机的关注。欧洲委员会已经拨款13亿美元用于这个项目。<br />
<br />
Other research with implications for neuromorphic engineering involves the [[BRAIN Initiative]]<ref name="economist">[https://www.economist.com/news/science-and-technology/21582495-computers-will-help-people-understand-brains-better-and-understanding-brains Neuromorphic computing: The machine of a new soul], The Economist, 2013-08-03</ref> and the [[TrueNorth]] chip from [[IBM]].<ref>{{cite journal|last1=Modha|first1=Dharmendra|title=A million spiking-neuron integrated circuit with a scalable communication network and interface|journal=Science|date=Aug 2014|volume=345|issue=6197|pages=668–673|doi=10.1126/science.1254642|pmid=25104385|bibcode=2014Sci...345..668M|s2cid=12706847}}</ref> Neuromorphic devices have also been demonstrated using nanocrystals, nanowires, and conducting polymers.<ref>{{Cite web|url=http://jessamynfairfield.com/wp-content/uploads/2017/03/PWMar17Fairfield.pdf|title=Smarter Machines|last=Fairfield|first=Jessamyn|date=March 1, 2017}}</ref><br />
<br />
Other research with implications for neuromorphic engineering involves the BRAIN InitiativeNeuromorphic computing: The machine of a new soul, The Economist, 2013-08-03 and the TrueNorth chip from IBM. Neuromorphic devices have also been demonstrated using nanocrystals, nanowires, and conducting polymers.<br />
<br />
其他与神经形态工程有关的研究还包括 BRAIN initiativity/euromorphic computing: The machine of a new soul，The Economist，2013-08-03 and The TrueNorth chip from IBM。神经形态学设备也已经被证明使用纳米晶体、纳米线和导电聚合物。<br />
<br />
[[Intel]] unveiled its neuromorphic research chip, called “[[Intel Loihi|Loihi]]”, in October 2017. The chip uses an asynchronous [[spiking neural network]] (SNN) to implement adaptive self-modifying event-driven fine-grained parallel computations used to implement learning and inference with high efficiency.<ref>{{cite journal |last1=Davies |first1=Mike |title=Loihi: A Neuromorphic Manycore Processor with On-Chip Learning |journal=IEEE Micro |date=January 16, 2018 |volume=38 |issue=1 |pages=82–99 |display-authors=etal|doi=10.1109/MM.2018.112130359 |s2cid=3608458 }}</ref><ref name="Morris2017">{{cite web |last1=Morris |first1=John |title=Why Intel built a neuromorphic chip |url=https://www.zdnet.com/article/why-intel-built-a-neuromorphic-chip/ |website=ZDNet |access-date=17 August 2018 |language=en}}</ref><br />
<br />
Intel unveiled its neuromorphic research chip, called “Loihi”, in October 2017. The chip uses an asynchronous spiking neural network (SNN) to implement adaptive self-modifying event-driven fine-grained parallel computations used to implement learning and inference with high efficiency.<br />
<br />
2017年10月，英特尔公布了其神经形态研究芯片“ Loihi”。该芯片采用异步脉冲神经网络(SNN)实现自适应自修改事件驱动的细粒度并行计算，实现了高效的学习和推理。<br />
<br />
[[IMEC]], a Belgium-based nanoelectronics research center, demonstrated the world's first self-learning neuromorphic chip. The brain-inspired chip, based on OxRAM technology, has the capability of self-learning and has been demonstrated to have the ability to compose music.<ref>{{cite web |title=Imec demonstrates self-learning neuromorphic chip that composes music |url=https://www.imec-int.com/en/articles/imec-demonstrates-self-learning-neuromorphic-chip-that-composes-music |website=IMEC International |access-date=1 October 2019}}</ref> IMEC released the 30-second tune composed by the prototype. The chip was sequentially loaded with songs in the same time signature and style. The songs were old Belgian and French flute minuets, from which the chip learned the rules at play and then applied them.<ref>{{cite web|last1=Bourzac|first1=Katherine|title=A Neuromorphic Chip That Makes Music|url=https://spectrum.ieee.org/a-neuromorphic-chip-that-makes-music|url-status=live|access-date=1 October 2019|website=IEEE Spectrum|date=May 23, 2017}}</ref><br />
<br />
IMEC, a Belgium-based nanoelectronics research center, demonstrated the world's first self-learning neuromorphic chip. The brain-inspired chip, based on OxRAM technology, has the capability of self-learning and has been demonstrated to have the ability to compose music. IMEC released the 30-second tune composed by the prototype. The chip was sequentially loaded with songs in the same time signature and style. The songs were old Belgian and French flute minuets, from which the chip learned the rules at play and then applied them.<br />
<br />
比利时的一个纳米电子学研究中心 IMEC 展示了世界上第一个自学习神经形态芯片。这种基于 OxRAM 技术的大脑启发芯片具有自学习能力，并已被证明具有创作音乐的能力。IMEC 发布了由原型机谱写的30秒曲调。芯片按顺序加载同时签名和风格的歌曲。歌曲是古老的比利时和法国长笛小步舞曲，筹码从中学习游戏规则，然后应用它们。<br />
<br />
[[Blue Brain Project|The Blue Brain Project]], led by Henry Markram, aims to build biologically detailed digital reconstructions and simulations of the mouse brain. The Blue Brain Project has created in silico models of rodent brains, while attempting to replicate as many details about its biology as possible. The supercomputer-based simulations offer new perspectives on understanding the structure and functions of the brain.<br />
<br />
The Blue Brain Project, led by Henry Markram, aims to build biologically detailed digital reconstructions and simulations of the mouse brain. The Blue Brain Project has created in silico models of rodent brains, while attempting to replicate as many details about its biology as possible. The supercomputer-based simulations offer new perspectives on understanding the structure and functions of the brain.<br />
<br />
由亨利 · 马克拉姆领导的蓝脑计划旨在建立小鼠大脑的生物细节数字重建和模拟。蓝色大脑项目已经在啮齿动物大脑的电子模型中创建，同时试图尽可能多地复制其生物学细节。基于超级计算机的模拟为理解大脑的结构和功能提供了新的视角。<br />
<br />
The European Union funded a series of projects at the University of Heidelberg, which led to the development of <br />
[[BrainScaleS]] (brain-inspired multiscale computation in neuromorphic hybrid systems), a hybrid analog [[neuromorphic]] supercomputer located at Heidelberg University, Germany. It was developed as part of the [[Human Brain Project]] neuromorphic computing platform and is the complement to the [[SpiNNaker]] supercomputer (which is based on digital technology). The architecture used in BrainScaleS mimics biological neurons and their connections on a physical level; additionally, since the components are made of silicon, these model neurons operate on average 864 times (24 hours of real time is 100 seconds in the machine simulation) that of their biological counterparts.<ref>{{Cite web|date=2016-03-21|title=Beyond von Neumann, Neuromorphic Computing Steadily Advances|url=https://www.hpcwire.com/2016/03/21/lacking-breakthrough-neuromorphic-computing-steadily-advance/|access-date=2021-10-08|website=HPCwire|language=en-US}}</ref><br />
<br />
The European Union funded a series of projects at the University of Heidelberg, which led to the development of <br />
BrainScaleS (brain-inspired multiscale computation in neuromorphic hybrid systems), a hybrid analog neuromorphic supercomputer located at Heidelberg University, Germany. It was developed as part of the Human Brain Project neuromorphic computing platform and is the complement to the SpiNNaker supercomputer (which is based on digital technology). The architecture used in BrainScaleS mimics biological neurons and their connections on a physical level; additionally, since the components are made of silicon, these model neurons operate on average 864 times (24 hours of real time is 100 seconds in the machine simulation) that of their biological counterparts.<br />
<br />
欧盟资助了海德堡大学的一系列项目，这些项目导致了 BrainScaleS (神经形态混合系统中受大脑启发的多尺度计算)的发展，这是一台位于德国海德堡大学的混合模拟神经形态超级计算机。它是作为人脑计划神经形态计算平台的一部分而开发的，是 SpiNNaker 超级计算机(基于数字技术)的补充。大脑尺度中使用的体系结构模拟了生物神经元及其在物理层面上的连接; 此外，由于这些组件是由硅制成的，这些模型神经元平均运行864倍(在机器模拟中，24小时的实时时间是100秒)。<br />
<br />
===Neuromorphic sensors===<br />
===神经形态传感器===<br />
The concept of neuromorphic systems can be extended to sensors (not just to computation). An example of this applied to detecting [[light]] is the [[retinomorphic sensor]] or, when employed in an array, the [[event camera]].<br />
<br />
The concept of neuromorphic systems can be extended to sensors (not just to computation). An example of this applied to detecting light is the retinomorphic sensor or, when employed in an array, the event camera.<br />
<br />
神经形态系统的概念可以扩展到传感器(而不仅仅是计算)。用于检测光线的一个例子是视网膜变形传感器，或者在阵列中使用的事件摄像机。<br />
<br />
==Ethical considerations ==<br />
==伦理问题==<br />
While the interdisciplinary concept of neuromorphic engineering is relatively new, many of the same ethical considerations apply to neuromorphic systems as apply to [[human-like machines]] and [[artificial intelligence]] in general. However, the fact that neuromorphic systems are designed to mimic a [[human brain]] gives rise to unique ethical questions surrounding their usage.<br />
<br />
While the interdisciplinary concept of neuromorphic engineering is relatively new, many of the same ethical considerations apply to neuromorphic systems as apply to human-like machines and artificial intelligence in general. However, the fact that neuromorphic systems are designed to mimic a human brain gives rise to unique ethical questions surrounding their usage.<br />
<br />
虽然神经形态工程的跨学科概念相对较新，但许多同样的伦理考虑适用于神经形态系统，就像适用于类人机器和一般人工智能一样。然而，神经形态系统是为了模仿人类大脑而设计的这一事实引起了围绕其使用的独特的伦理问题。<br />
<br />
However, the practical debate is that neuromorphic hardware as well as artificial "neural networks" are immensely simplified models of how the brain operates or processes information at a much lower [[complex system|complexity in terms of size and functional technology]] and a much more regular structure in terms of [[brain connectivity|connectivity]]. Comparing [[neuromorphic chip]]s to the brain is a very crude comparison similar to comparing a plane to a bird just because they both have wings and a tail. The fact is that neural cognitive systems are many orders of magnitude more [[Energy efficiency (physics)|energy-]] and compute-efficient than current state-of-the-art AI and neuromorphic engineering is an attempt to narrow this gap by inspiring from the brain's mechanism just like many engineering designs have [[bioengineering|bio-inspired features]].<br />
<br />
However, the practical debate is that neuromorphic hardware as well as artificial "neural networks" are immensely simplified models of how the brain operates or processes information at a much lower complexity in terms of size and functional technology and a much more regular structure in terms of connectivity. Comparing neuromorphic chips to the brain is a very crude comparison similar to comparing a plane to a bird just because they both have wings and a tail. The fact is that neural cognitive systems are many orders of magnitude more energy- and compute-efficient than current state-of-the-art AI and neuromorphic engineering is an attempt to narrow this gap by inspiring from the brain's mechanism just like many engineering designs have bio-inspired features.<br />
<br />
然而，实际上的争论是，神经形态硬件和人工“神经网络”是大脑如何运作或处理信息的极其简化的模型，在大小和功能技术方面的复杂性要低得多，在连接方面的结构则更加规则。将神经形态芯片与大脑进行比较是一种非常粗糙的比较，类似于仅仅因为一架飞机有翅膀和一条尾巴就将它与一只鸟进行比较。事实上，神经认知系统比当前最先进的人工智能具有更多的能量和计算效率，而神经形态工程是一种通过从大脑机制中激发灵感来缩小这种差距的尝试，就像许多工程设计具有生物启发的特征一样。数量级。<br />
<br />
===Democratic concerns===<br />
===公众担忧===<br />
Significant ethical limitations may be placed on neuromorphic engineering due to public perception.<ref>{{Cite report|url=https://ai100.stanford.edu/sites/g/files/sbiybj9861/f/ai_100_report_0831fnl.pdf|title=Artificial Intelligence and Life in 2030|author=2015 Study Panel|date=September 2016|work=One Hundred Year Study on Artificial Intelligence (AI100)|publisher=Stanford University}}</ref> Special [[Eurobarometer]] 382: Public Attitudes Towards Robots, a survey conducted by the European Commission, found that 60% of [[European Union]] citizens wanted a ban of robots in the care of children, the elderly, or the disabled. Furthermore, 34% were in favor of a ban on robots in education, 27% in healthcare, and 20% in leisure. The European Commission classifies these areas as notably “human.” The report cites increased public concern with robots that are able to mimic or replicate human functions. Neuromorphic engineering, by definition, is designed to replicate the function of the human brain.<ref name=":1">{{Cite web|url=http://ec.europa.eu/commfrontoffice/publicopinion/archives/ebs/ebs_382_en.pdf|title=Special Eurobarometer 382: Public Attitudes Towards Robots|last=European Commission|date=September 2012|website=European Commission}}</ref><br />
<br />
Significant ethical limitations may be placed on neuromorphic engineering due to public perception. Special Eurobarometer 382: Public Attitudes Towards Robots, a survey conducted by the European Commission, found that 60% of European Union citizens wanted a ban of robots in the care of children, the elderly, or the disabled. Furthermore, 34% were in favor of a ban on robots in education, 27% in healthcare, and 20% in leisure. The European Commission classifies these areas as notably “human.” The report cites increased public concern with robots that are able to mimic or replicate human functions. Neuromorphic engineering, by definition, is designed to replicate the function of the human brain.<br />
<br />
由于公众的认知，神经形态工程学可能会受到严重的伦理限制。欧盟委员会进行的一项调查发现，60% 的欧盟公民希望禁止照顾儿童、老人或残疾人的机器人。此外，34% 的人支持禁止机器人用于教育，27% 的人支持医疗保健，20% 的人支持休闲。欧盟委员会将这些地区明显归类为“人类”报告指出，公众越来越关注能够模仿或复制人类功能的机器人。神经形态工程，顾名思义，是为了复制人脑的功能而设计的。<br />
<br />
The democratic concerns surrounding neuromorphic engineering are likely to become even more profound in the future. The European Commission found that EU citizens between the ages of 15 and 24 are more likely to think of robots as human-like (as opposed to instrument-like) than EU citizens over the age of 55. When presented an image of a robot that had been defined as human-like, 75% of EU citizens aged 15–24 said it corresponded with the idea they had of robots while only 57% of EU citizens over the age of 55 responded the same way. The human-like nature of neuromorphic systems, therefore, could place them in the categories of robots many EU citizens would like to see banned in the future.<ref name=":1" /><br />
<br />
The democratic concerns surrounding neuromorphic engineering are likely to become even more profound in the future. The European Commission found that EU citizens between the ages of 15 and 24 are more likely to think of robots as human-like (as opposed to instrument-like) than EU citizens over the age of 55. When presented an image of a robot that had been defined as human-like, 75% of EU citizens aged 15–24 said it corresponded with the idea they had of robots while only 57% of EU citizens over the age of 55 responded the same way. The human-like nature of neuromorphic systems, therefore, could place them in the categories of robots many EU citizens would like to see banned in the future.<br />
<br />
围绕神经形态工程的民主关注可能在未来变得更加深刻。欧盟委员会(European Commission)发现，15至24岁的欧盟公民比55岁以上的欧盟公民更有可能认为机器人像人(而不是像仪器)。当看到一张被定义为“类人”的机器人图片时，年龄在15岁至24岁之间的欧盟公民中有75% 的人表示，这与他们对机器人的想法相符，而55岁以上的欧盟公民中只有57% 的人有同样的反应。因此，类似人类的神经形态系统，可以把它们归入许多欧盟公民希望在未来禁止使用的机器人类别。<br />
<br />
===Personhood===<br />
===人格问题===<br />
As neuromorphic systems have become increasingly advanced, some scholars{{who|date=August 2021}} have advocated for granting [[personhood]] rights to these systems. If the brain is what grants humans their personhood, to what extent does a neuromorphic system have to mimic the human brain to be granted personhood rights? Critics of technology development in the [[Human Brain Project]], which aims to advance brain-inspired computing, have argued that advancement in neuromorphic computing could lead to machine consciousness or personhood.<ref>{{Cite journal|last=Aicardi|first=Christine|date=September 2018|title=Accompanying technology development in the Human Brain Project: From foresight to ethics management|journal=Futures|volume=102|pages=114–124|doi=10.1016/j.futures.2018.01.005|doi-access=free}}</ref> If these systems are to be treated as people, critics argue, then many tasks humans perform using neuromorphic systems, including the act of termination of neuromorphic systems, may be morally impermissible as these acts would violate the autonomy of the neuromorphic systems.<ref>{{Cite journal|last=Lim|first=Daniel|date=2014-06-01|title=Brain simulation and personhood: a concern with the Human Brain Project|journal=Ethics and Information Technology|language=en|volume=16|issue=2|pages=77–89|doi=10.1007/s10676-013-9330-5|s2cid=17415814|issn=1572-8439}}</ref><br />
<br />
As neuromorphic systems have become increasingly advanced, some scholars have advocated for granting personhood rights to these systems. If the brain is what grants humans their personhood, to what extent does a neuromorphic system have to mimic the human brain to be granted personhood rights? Critics of technology development in the Human Brain Project, which aims to advance brain-inspired computing, have argued that advancement in neuromorphic computing could lead to machine consciousness or personhood. If these systems are to be treated as people, critics argue, then many tasks humans perform using neuromorphic systems, including the act of termination of neuromorphic systems, may be morally impermissible as these acts would violate the autonomy of the neuromorphic systems.<br />
<br />
随着神经形态系统的日益发展，一些学者主张赋予这些系统人格权。如果是大脑赋予了人类人格，那么神经形态系统在多大程度上必须模仿人类大脑才能被赋予人格权利？“人脑计划”旨在推进以大脑为灵感的计算机技术发展，该计划的批评者认为，神经形态计算机技术的进步可能导致机器意识或人格的形成。批评家认为，如果这些系统被当作人来对待，那么人类使用神经形态系统执行的许多任务，包括终止神经形态系统的行为，可能在道德上是不允许的，因为这些行为将违反神经形态系统的自主性。<br />
<br />
==Dual use (military applications)==<br />
The [[Joint Artificial Intelligence Center]], a branch of the U.S. military, is a center dedicated to the procurement and implementation of AI software and neuromorphic hardware for combat use. Specific applications include smart headsets/goggles and robots. JAIC intends to rely heavily on neuromorphic technology to connect "every fighter every shooter" within a network of neuromorphic-enabled units.<br />
<br />
The Joint Artificial Intelligence Center, a branch of the U.S. military, is a center dedicated to the procurement and implementation of AI software and neuromorphic hardware for combat use. Specific applications include smart headsets/goggles and robots. JAIC intends to rely heavily on neuromorphic technology to connect "every fighter every shooter" within a network of neuromorphic-enabled units.<br />
<br />
军民两用联合人工智能中心是美国军队的一个分支，专门从事采购和实施用于战斗的人工智能软件和神经形态硬件。具体应用包括智能耳机/护目镜和机器人。JAIC 打算严重依赖神经形态技术来连接神经形态单位网络中的“每个战士每个射手”。<br />
<br />
==Legal considerations ==<br />
==法律问题==<br />
Skeptics have argued that there is no way to apply the electronic personhood, the concept of personhood that would apply to neuromorphic technology, legally. In a letter signed by 285 experts in law, robotics, medicine, and ethics opposing a European Commission proposal to recognize “smart robots” as legal persons, the authors write, “A legal status for a robot can’t derive from the [[Natural person|Natural Person]] model, since the robot would then hold [[human rights]], such as the right to dignity, the right to its integrity, the right to remuneration or the right to citizenship, thus directly confronting the Human rights. This would be in contradiction with the [[Charter of Fundamental Rights of the European Union]] and the [[Convention for the Protection of Human Rights and Fundamental Freedoms]].”<ref>{{Cite web|url=http://www.robotics-openletter.eu/|title=Robotics Openletter {{!}} Open letter to the European Commission|language=fr-FR|access-date=2019-05-10}}</ref><br />
<br />
Skeptics have argued that there is no way to apply the electronic personhood, the concept of personhood that would apply to neuromorphic technology, legally. In a letter signed by 285 experts in law, robotics, medicine, and ethics opposing a European Commission proposal to recognize “smart robots” as legal persons, the authors write, “A legal status for a robot can’t derive from the Natural Person model, since the robot would then hold human rights, such as the right to dignity, the right to its integrity, the right to remuneration or the right to citizenship, thus directly confronting the Human rights. This would be in contradiction with the Charter of Fundamental Rights of the European Union and the Convention for the Protection of Human Rights and Fundamental Freedoms.”<br />
<br />
怀疑论者认为，在法律上没有办法应用电子人格，这个人格概念将适用于神经形态技术。在一封由285名法律、机器人、医学和伦理学专家签名的信中，作者们反对欧盟委员会承认“智能机器人”为法人的提议，他们写道，“机器人的法律地位不能从自然人模型中推导出来，因为机器人将拥有人权，如尊严权、完整权、报酬权或公民权，从而直接面对人权。这将有悖于《欧洲联盟基本权利宪章和《保护人权和基本自由公约》。”<br />
<br />
===Ownership and property rights===<br />
===所有权及财产权问题 ===<br />
There is significant legal debate around property rights and artificial intelligence. In ''Acohs Pty Ltd v. Ucorp Pty Ltd'', Justice Christopher Jessup of the [[Federal Court of Australia]] found that the [[source code]] for [[Material safety data sheets|Material Safety Data Sheets]] could not be [[Copyright law of Australia|copyrighted]] as it was generated by a [[software interface]] rather than a human author.<ref>{{Cite web|url=http://www.lavan.com.au/advice/intellectual_property_technology/copyright_in_source_code_and_digital_products|title=Copyright in source code and digital products|last=Lavan|website=Lavan|language=en|access-date=2019-05-10}}</ref> The same question may apply to neuromorphic systems: if a neuromorphic system successfully mimics a human brain and produces a piece of original work, who, if anyone, should be able to claim ownership of the work?<ref>{{cite journal |last1=Eshraghian|first1=Jason K. |title=Human Ownership of Artificial Creativity |journal=Nature Machine Intelligence |date=9 March 2020 |volume=2 |pages=157–160 |doi=10.1038/s42256-020-0161-x}}</ref><br />
<br />
There is significant legal debate around property rights and artificial intelligence. In Acohs Pty Ltd v. Ucorp Pty Ltd, Justice Christopher Jessup of the Federal Court of Australia found that the source code for Material Safety Data Sheets could not be copyrighted as it was generated by a software interface rather than a human author. The same question may apply to neuromorphic systems: if a neuromorphic system successfully mimics a human brain and produces a piece of original work, who, if anyone, should be able to claim ownership of the work?<br />
<br />
围绕财产权和人工智能有着重大的法律争论。在 Acohs Pty Ltd 诉 Ucorp Pty Ltd 一案中，澳大利亚联邦法院的克里斯托弗 · 杰瑟普法官发现，材料安全数据表的源代码不能受版权保护，因为它是由软件界面而不是人工作者生成的。同样的问题可能也适用于神经形态系统: 如果一个神经形态系统成功地模仿了人类的大脑并产生了一部原创作品，那么谁，如果有人，应该声称拥有这部作品的所有权？<br />
<br />
==Neuromemristive systems==<br />
==神经记忆电阻系统==<br />
Neuromemristive systems are a subclass of neuromorphic computing systems that focus on the use of [[memristors]] to implement [[neuroplasticity]]. While neuromorphic engineering focuses on mimicking biological behavior, neuromemristive systems focus on abstraction.<ref>{{Cite web|url=https://digitalops.sandia.gov/Mediasite/Play/a10cf6ceb55d47608bb8326dd00e46611d|title=002.08 N.I.C.E. Workshop 2014: Towards Intelligent Computing with Neuromemristive Circuits and Systems - Feb. 2014|website=digitalops.sandia.gov|access-date=2019-08-26}}</ref> For example, a neuromemristive system may replace the details of a [[Cerebral cortex|cortical]] microcircuit's behavior with an abstract neural network model.<ref>C. Merkel and D. Kudithipudi, "Neuromemristive extreme learning machines for pattern classification," ISVLSI, 2014.</ref><br />
<br />
Neuromemristive systems are a subclass of neuromorphic computing systems that focus on the use of memristors to implement neuroplasticity. While neuromorphic engineering focuses on mimicking biological behavior, neuromemristive systems focus on abstraction. For example, a neuromemristive system may replace the details of a cortical microcircuit's behavior with an abstract neural network model.C. Merkel and D. Kudithipudi, "Neuromemristive extreme learning machines for pattern classification," ISVLSI, 2014.<br />
<br />
神经记忆电阻系统是神经形态计算系统的一个亚类，主要研究利用记忆电阻器实现神经可塑性。神经形态工程的重点是模拟生物行为，而神经记忆电阻系统的重点是提取。例如，一个神经记忆系统可以用一个抽象的神经网络模型取代皮层微电路的行为细节。默克尔和 d. Kudithipudi，“用于模式分类的神经记忆极端学习机器”，ISVLSI，2014。<br />
<br />
There exist several neuron inspired threshold logic functions<ref name="Maan 1–13" /> implemented with memristors that have applications in high level [[pattern recognition]] applications. Some of the applications reported recently include [[speech recognition]],<ref>{{Cite journal|title = Memristor pattern recogniser: isolated speech word recognition|journal = Electronics Letters|pages = 1370–1372|volume = 51|issue = 17|doi = 10.1049/el.2015.1428|first1 = A.K.|last1 = Maan|first2 = A.P.|last2 = James|first3 = S.|last3 = Dimitrijev|year = 2015|bibcode = 2015ElL....51.1370M|hdl = 10072/140989|s2cid = 61454815|url = https://semanticscholar.org/paper/48d3ab11ec6e213b62f11eedcfb7b7febb058674|hdl-access = free}}</ref> [[face recognition]]<ref>{{Cite journal|title = Memristive Threshold Logic Face Recognition|journal = Procedia Computer Science|date = 2014-01-01|pages = 98–103|volume = 41|series = 5th Annual International Conference on Biologically Inspired Cognitive Architectures, 2014 BICA|doi = 10.1016/j.procs.2014.11.090|first1 = Akshay Kumar|last1 = Maan|first2 = Dinesh S.|last2 = Kumar|first3 = Alex Pappachen|last3 = James|doi-access = free}}</ref> and [[object recognition]].<ref>{{Cite journal|title = Memristive Threshold Logic Circuit Design of Fast Moving Object Detection|journal = IEEE Transactions on Very Large Scale Integration (VLSI) Systems|date = 2015-10-01|issn = 1063-8210|pages = 2337–2341|volume = 23|issue = 10|doi = 10.1109/TVLSI.2014.2359801|first1 = A.K.|last1 = Maan|first2 = D.S.|last2 = Kumar|first3 = S.|last3 = Sugathan|first4 = A.P.|last4 = James|arxiv = 1410.1267|s2cid = 9647290}}</ref> They also find applications in replacing conventional digital logic gates.<ref>{{Cite journal|title = Resistive Threshold Logic|journal = IEEE Transactions on Very Large Scale Integration (VLSI) Systems|date = 2014-01-01|issn = 1063-8210|pages = 190–195|volume = 22|issue = 1|doi = 10.1109/TVLSI.2012.2232946|first1 = A.P.|last1 = James|first2 = L.R.V.J.|last2 = Francis|first3 = D.S.|last3 = Kumar|arxiv = 1308.0090|s2cid = 7357110}}</ref><ref>{{Cite journal|title = Threshold Logic Computing: Memristive-CMOS Circuits for Fast Fourier Transform and Vedic Multiplication|journal = IEEE Transactions on Very Large Scale Integration (VLSI) Systems|date = 2015-11-01|issn = 1063-8210|pages = 2690–2694|volume = 23|issue = 11|doi = 10.1109/TVLSI.2014.2371857|first1 = A.P.|last1 = James|first2 = D.S.|last2 = Kumar|first3 = A.|last3 = Ajayan|arxiv = 1411.5255|s2cid = 6076956}}</ref><br />
<br />
There exist several neuron inspired threshold logic functions implemented with memristors that have applications in high level pattern recognition applications. Some of the applications reported recently include speech recognition, face recognition and object recognition. They also find applications in replacing conventional digital logic gates.<br />
<br />
利用记忆电阻器实现的神经元启发阈值逻辑函数在高级模式识别中有着广泛的应用。最近报道的一些应用包括语音识别、人脸识别和物体识别。它们还可以用来取代传统的数字逻辑门。<br />
<br />
For ideal passive memristive circuits there is an exact equation (Caravelli-Traversa-[[Di Ventra]] equation) for the internal memory of the circuit:<ref>{{cite journal |last=Caravelli |display-authors=etal|arxiv=1608.08651 |title=The complex dynamics of memristive circuits: analytical results and universal slow relaxation |year=2017 |doi=10.1103/PhysRevE.95.022140 |pmid= 28297937 |volume=95 |issue= 2 |pages= 022140 |journal=Physical Review E|bibcode=2017PhRvE..95b2140C |s2cid=6758362}}</ref><br />
<br />
For ideal passive memristive circuits there is an exact equation (Caravelli-Traversa-Di Ventra equation) for the internal memory of the circuit:<br />
<br />
对于理想的无源记忆电路，电路的内部记忆有一个精确的方程(Caravelli-Traversa-Di Ventra 方程) :<br />
<br />
:<math> \frac{d}{dt} \vec{W} = \alpha \vec{W}-\frac{1}{\beta} (I+\xi \Omega W)^{-1} \Omega \vec S </math><br />
<br />
:\frac{d}{dt} \vec{W} = \alpha \vec{W}-\frac{1}{\beta} (I+\xi \Omega W)^{-1} \Omega \vec S<br />
<br />
:\frac{d}{dt} \vec{W} = \alpha \vec{W}-\frac{1}{\beta} (I+\xi \Omega W)^{-1} \Omega \vec S<br />
<br />
as a function of the properties of the physical memristive network and the external sources. In the equation above, <math>\alpha</math> is the "forgetting" time scale constant, <math> \xi=r-1</math> and <math>r=\frac{R_\text{off}}{R_\text{on}}</math> is the ratio of ''off'' and ''on'' values of the limit resistances of the memristors, <math> \vec S </math> is the vector of the sources of the circuit and <math>\Omega</math> is a projector on the fundamental loops of the circuit. The constant <math>\beta</math> has the dimension of a voltage and is associated to the properties of the [[memristor]]; its physical origin is the charge mobility in the conductor. The diagonal matrix and vector <math>W=\operatorname{diag}(\vec W)</math> and <math>\vec W</math> respectively, are instead the internal value of the memristors, with values between 0 and 1. This equation thus requires adding extra constraints on the memory values in order to be reliable.<br />
<br />
<nowiki>as a function of the properties of the physical memristive network and the external sources. In the equation above, \alpha is the "forgetting" time scale constant, \xi=r-1 and r=\frac{R_\text{off}}{R_\text{on}} is the ratio of off and on values of the limit resistances of the memristors, \vec S is the vector of the sources of the circuit and \Omega is a projector on the fundamental loops of the circuit. The constant \beta has the dimension of a voltage and is associated to the properties of the memristor; its physical origin is the charge mobility in the conductor. The diagonal matrix and vector W=\operatorname{diag}(\vec W) and \vec W respectively, are instead the internal value of the memristors, with values between 0 and 1. This equation thus requires adding extra constraints on the memory values in order to be reliable.</nowiki><br />
<br />
<nowiki>作为物理记忆网络和外部源的性质的函数。在上述方程中，α 是“遗忘”时间尺度常数，xi = r-1，r = frac { r _ text { off }{ on }{ r _ text { on }}是记忆电阻器极限电阻的开关和开关值之比，vec s 是电路源的矢量，Omega 是电路基本环路的投影仪。常数 β 具有电压的尺寸，与记忆电阻器的特性有关; 它的物理起源是导体中的电荷迁移率。对角矩阵和向量 w = 操作者名{ diag }(vec w)和 vec w 分别是记忆电阻器的内值，值在0到1之间。因此，这个等式需要在内存值上添加额外的约束，以保证可靠性。</nowiki><br />
<br />
==See also==<br />
== 相关词条==<br />
{{Columns-list|colwidth=18em|<br />
* [[AI accelerator (computer hardware)|AI accelerator]]<br />
* [[Artificial brain]]<br />
* [[Biomorphic]]<br />
* [[Cognitive computer]]<br />
* [[Computation and Neural Systems]]<br />
* [[Differentiable programming]]<br />
* [[Event camera]]<br />
* [[Neurorobotics]]<br />
* [[Optical flow sensor]]<br />
* [[Physical neural network]]<br />
* [[SpiNNaker]]<br />
* [[SyNAPSE]]<br />
* [[Retinomorphic sensor]]<br />
* [[Vision chip]]<br />
* [[Vision processing unit]]<br />
* [[Zeroth (software)]]<br />
* [[Hardware for artificial intelligence]]<br />
}}<br />
<br />
{{Portal bar|Electronics}}<br />
<br />
==References==<br />
{{Reflist|40em}}<br />
<br />
==External links==<br />
<!--======================== {{No more links}} ============================<br />
| PLEASE BE CAUTIOUS IN ADDING MORE LINKS TO THIS ARTICLE. Wikipedia |<br />
| is not a collection of links nor should it be used for advertising. |<br />
| |<br />
| Excessive or inappropriate links WILL BE DELETED. |<br />
| See [[Wikipedia:External links]] & [[Wikipedia:Spam]] for details. |<br />
| |<br />
| If there are already plentiful links, please propose additions or |<br />
| replacements on this article's discussion page, or submit your link |<br />
| to the relevant category at the Open Directory Project (dmoz.org) |<br />
| and link back to that category using the {{dmoz}} template. |<br />
======================={{No more links}}=============================--><br />
{{Commons category|Neuromorphic Engineering}}<br />
*[https://web.archive.org/web/20150727034331/http://ine-web.org/workshops/workshops-overview Telluride Neuromorphic Engineering Workshop]<br />
*[https://archive.today/20130115190057/http://capocaccia.ethz.ch/ CapoCaccia Cognitive Neuromorphic Engineering Workshop]<br />
*[https://web.archive.org/web/20190716132350/http://www.ine-web.org/ Institute of Neuromorphic Engineering]<br />
*[http://www.ine-news.org/ INE news site].<br />
*[http://www.frontiersin.org/neuromorphic_engineering Frontiers in Neuromorphic Engineering Journal]<br />
*[http://www.cns.caltech.edu/ Computation and Neural Systems] department at the [[California Institute of Technology]].<br />
*[http://www.humanbrainproject.eu/ Human Brain Project official site]<br />
*[https://www.the-scientist.com/features/building-a-silicon-brain-65738 Building a Silicon Brain:] Computer chips based on biological neurons may help simulate larger and more-complex brain models. May 1, 2019. SANDEEP RAVINDRAN<br />
<br />
<br />
*Telluride Neuromorphic Engineering Workshop<br />
*CapoCaccia Cognitive Neuromorphic Engineering Workshop<br />
*Institute of Neuromorphic Engineering<br />
*INE news site.<br />
*Frontiers in Neuromorphic Engineering Journal<br />
*Computation and Neural Systems department at the California Institute of Technology.<br />
*Human Brain Project official site<br />
*Building a Silicon Brain: Computer chips based on biological neurons may help simulate larger and more-complex brain models. May 1, 2019. SANDEEP RAVINDRAN<br />
<br />
=<nowiki>= = 外部链接 =</nowiki>= <br />
*碲化物神经形态工程工作室<br />
*CapoCaccia 认知神经形态工程工作室<br />
*神经形态工程研究所<br />
*INE 新闻站点。<br />
*《神经形态工程学前沿》<br />
*加州理工学院计算与神经系统系。<br />
*人脑项目官方网站<br />
*打造硅脑: 基于生物神经元的计算机芯片可能有助于模拟更大、更复杂的大脑模型。2019年5月1日。SANDEEP RAVINDRAN<br />
<br />
{{Differentiable computing}}<br />
{{Authority control}}<br />
<br />
[[Category:Electrical engineering]]<br />
[[Category:Neuroscience]]<br />
[[分类:AI accelerators|*]]<br />
[[Category:Artificial intelligence]]<br />
[[Category:Robotics]]<br />
<br />
Category:Electrical engineering<br />
Category:Neuroscience<br />
*<br />
Category:Artificial intelligence<br />
Category:Robotics<br />
<br />
类别: 电气工程类别: 神经科学 * 类别: 人工智能类别: 机器人<br />
<br />
<noinclude><br />
<br />
<small>This page was moved from [[wikipedia:en:Neuromorphic engineering]]. Its edit history can be viewed at [[神经形态计算/edithistory]]</small></noinclude><br />
<br />
[[Category:待整理页面]]<br />
<br />
<references /></div>Shenky20https://wiki.swarma.org/index.php?title=%E7%A5%9E%E7%BB%8F%E5%BD%A2%E6%80%81%E8%AE%A1%E7%AE%97&diff=30174神经形态计算2022-04-13T07:40:40Z<p>Shenky20：</p>
<hr />
<div><br />
此词条由神经动力学模型读书会词条梳理志愿者Shenky20翻译审校，未经专家审校，带来阅读不便，请见谅。<br />
<br />
{{Use American English|date = January 2019}}<br />
{{Use mdy dates|date = January 2019}}<br />
<br />
'''Neuromorphic engineering''', also known as '''neuromorphic computing''',<ref name=":3">{{Cite journal | doi = 10.1145/2601069| title = Neuromorphic computing gets ready for the (really) big time| journal = [[Communications of the ACM]]| volume = 57| issue = 6| year = 2014| last1 = Monroe | first1 = D. | pages = 13–15| s2cid = 20051102}}</ref><ref name=":4">{{Cite journal | doi = 10.1088/0957-4484/21/17/175202| pmid = 20368686| title = Nanotube devices based crossbar architecture: Toward neuromorphic computing| journal = Nanotechnology| volume = 21| issue = 17| pages = 175202| year = 2010| last1 = Zhao | first1 = W. S. | last2 = Agnus | first2 = G. | last3 = Derycke | first3 = V. | last4 = Filoramo | first4 = A. | last5 = Bourgoin | first5 = J. -P. | last6 = Gamrat | first6 = C. | bibcode = 2010Nanot..21q5202Z| url = https://zenodo.org/record/3428659}}</ref><ref name="humanbrainproject">{{YouTube|id=6RoiZ90mGfw|title=The Human Brain Project SP 9: Neuromorphic Computing Platform}}</ref> is the use of [[very-large-scale integration]] (VLSI) systems containing electronic [[analog circuit]]s to mimic neuro-biological architectures present in the nervous system. A neuromorphic computer/chip is any device that uses physical artificial neurons (made from silicon) to do computations.<ref name=":5">{{cite journal|last1=Mead|first1=Carver|title=Neuromorphic electronic systems|journal=Proceedings of the IEEE|date=1990|volume=78|issue=10|pages=1629–1636|doi=10.1109/5.58356|url=https://authors.library.caltech.edu/53090/1/00058356.pdf}}</ref><ref name=":2" /> In recent times, the term ''neuromorphic'' has been used to describe analog, digital, mixed-mode analog/digital VLSI, and software systems that implement models of [[neural system]]s (for [[perception]], [[motor control]], or [[multisensory integration]]). The implementation of neuromorphic computing on the hardware level can be realized by oxide-based [[memristor]]s,<ref name="Maan 1–13">{{Cite journal|last1=Maan|first1=A. K.|last2=Jayadevi|first2=D. A.|last3=James|first3=A. P.|date=2016-01-01|title=A Survey of Memristive Threshold Logic Circuits|journal=IEEE Transactions on Neural Networks and Learning Systems|volume=PP|issue=99|pages=1734–1746|doi=10.1109/TNNLS.2016.2547842|pmid=27164608|issn=2162-237X|arxiv=1604.07121|bibcode=2016arXiv160407121M|s2cid=1798273}}</ref> spintronic memories, threshold switches, and [[transistor]]s.<ref name=":6">{{Cite journal|title = Mott Memory and Neuromorphic Devices|journal = Proceedings of the IEEE|date = 2015-08-01|issn = 0018-9219|pages = 1289–1310|volume = 103|issue = 8|doi = 10.1109/JPROC.2015.2431914|first1 = You|last1 = Zhou|first2 = S.|last2 = Ramanathan|s2cid = 11347598|url = https://zenodo.org/record/895565}}</ref><ref name=":2">{{Cite document|title=Neuromorphic Circuits With Neural Modulation Enhancing the Information Content of Neural Signaling {{!}} International Conference on Neuromorphic Systems 2020|language=EN|doi=10.1145/3407197.3407204|s2cid=220794387}}</ref> Training software-based neuromorphic systems of spiking neural networks can be achieved using error backpropagation, e.g., using [[Python (programming language)|Python]] based frameworks such as snnTorch,<ref name=":7">{{cite journal |last1=Eshraghian|first1=Jason K.|last2=Ward|first2=Max|last3=Neftci |first3=Emre|last4=Wang|first4=Xinxin|last5=Lenz|first5=Gregor|last6=Dwivedi|first6=Girish|last7=Bennamoun|first7=Mohammed|last8=Jeong|first8=Doo Seok|last9=Lu|first9=Wei D.|title=Training Spiking Neural Networks Using Lessons from Deep Learning |date=1 October 2021 |arxiv=2109.12894 }}</ref> or using canonical learning rules from the biological learning literature, e.g., using BindsNet.<ref name=":8">{{Cite web | url=https://github.com/Hananel-Hazan/bindsnet | title=Hananel-Hazan/bindsnet: Simulation of spiking neural networks (SNNs) using PyTorch.| date=31 March 2020}}</ref><br />
<br />
<br />
'''<font color="#ff8000">神经形态工程Neuromorphic engineering</font>'''(也称为'''<font color="#ff8000">神经形态计算Neuromorphic computing</font>''')<ref name=":3" /><ref name=":4" /><ref name="humanbrainproject" />是指使用包含电子'''<font color="#ff8000">模拟电路Analog circuit</font>'''的'''<font color="#ff8000">超大规模集成电路Very-large-scale integration</font>'''系统来模拟神经系统中生理结构的研究方法。神经形态计算机或神经形态芯片包括任何使用由硅制成的人造神经元进行计算的设备。<ref name=":5" /><ref name=":2" />近年来，神经形态学（neuromorphic）这个术语被用来描述能够实现'''<font color="#ff8000">神经系统Neural system</font>'''模型功能(如'''<font color="#ff8000">感知Perception</font>'''、'''<font color="#ff8000">运动控制Motor control</font>'''，'''<font color="#ff8000">多感官整合Multisensory integration</font>'''等)的模拟、数字、模拟/数字混合模式超大规模集成电路和软件系统。神经形态计算的硬件实现可以通过基于氧化物的'''<font color="#ff8000">记忆电阻器Memristor</font>'''、<ref name="Maan 1–13" />自旋电子存储器、阈值开关和'''<font color="#ff8000">晶体管Transistor</font>'''来实现。<ref name=":6" /><ref name=":2" />对基于软件的脉冲神经网络系统的训练可以通过误差反向传播机制来实现，例如，使用snnTorch等基于Python的框架，<ref name=":7" />或使用BindsNet等典型的受生物启发的学习模式。<ref name=":8" /><br />
<br />
A key aspect of neuromorphic engineering is understanding how the morphology of individual neurons, circuits, applications, and overall architectures creates desirable computations, affects how information is represented, influences robustness to damage, incorporates learning and development, adapts to local change (plasticity), and facilitates evolutionary change.<br />
<br />
<br />
神经形态工程领域的一个关键问题，就是理解'''<font color="#32CD32">单个神经元形态、神经回路、应用和整体结构</font>'''如何产生理想的计算，如何影响信息的表示和对破坏的鲁棒性，如何整合学习和发展，如何适应局部变化(可塑性) 并促进逐渐发展的变化。<br />
<br />
Neuromorphic engineering is an interdisciplinary subject that takes inspiration from [[biology]], [[physics]], [[mathematics]], [[computer science]], and [[electronic engineering]]<ref name=":2" /> to design artificial neural systems, such as [[Machine vision|vision systems]], head-eye systems, auditory processors, and autonomous robots, whose physical architecture and design principles are based on those of biological nervous systems.<ref name=":9">{{Cite journal | doi = 10.1155/2012/705483| title = Qualitative Functional Decomposition Analysis of Evolved Neuromorphic Flight Controllers| journal = Applied Computational Intelligence and Soft Computing| volume = 2012| pages = 1–21| year = 2012| last1 = Boddhu | first1 = S. K. | last2 = Gallagher | first2 = J. C. | doi-access = free}}</ref> It was developed by [[Carver Mead]]<ref name=":10">{{cite web|title=carver mead website|last1=Mead|first1=Carver|url=http://www.carvermead.caltech.edu/index.html|website=carvermead}}</ref> in the late 1980s.<br />
<br />
<br />
神经形态工程是以'''<font color="#ff8000">生物学Biology</font>'''、'''<font color="#ff8000">物理学Physics</font>'''、'''<font color="#ff8000">数学Mathematics</font>'''、'''<font color="#ff8000">计算机科学Computer science</font>'''和'''<font color="#ff8000">电子工程Electronic engineering<ref name=":2" /></font>'''等学科为基础，设计人工神经系统（如'''<font color="#ff8000">视觉系统Vision systems</font>'''、头眼系统、听觉处理器以及物理结构和设计原则都受启发于生物神经系统的自主机器人）的一门交叉学科。<ref name=":9" />20世纪80年代后期，卡弗·米德极大地推动了神经形态工程领域的发展。<ref name=":10" /><br />
<br />
==Neurological inspiration==<br />
Neuromorphic engineering is set apart by the inspiration it takes from what we know about the structure and operations of the [[brain]]. Neuromorphic engineering translates what we know about the brain's function into computer systems. Work has mostly focused on replicating the analog nature of [[biological computation]] and the role of [[neuron]]s in [[cognition]].<br />
<br />
Neuromorphic engineering is set apart by the inspiration it takes from what we know about the structure and operations of the brain. Neuromorphic engineering translates what we know about the brain's function into computer systems. Work has mostly focused on replicating the analog nature of biological computation and the role of neurons in cognition.<br />
<br />
神经学灵感工程学是从我们对大脑结构和运作的了解中获得灵感而形成的。神经形态工程学将我们对大脑功能的了解转化为计算机系统。工作主要集中在复制生物计算的模拟特性和神经元在认知中的作用。<br />
<br />
The biological processes of neurons and their [[synapse]]s are dauntingly complex, and thus very difficult to artificially simulate. A key feature of biological brains is that all of the processing in neurons use analog chemical signals. This makes it hard to replicate brains in computers because the current generation of computers is completely digital. However, the characteristics of these parts can be abstracted into mathematical functions that closely capture the essence of the neuron's operations.<br />
<br />
The biological processes of neurons and their synapses are dauntingly complex, and thus very difficult to artificially simulate. A key feature of biological brains is that all of the processing in neurons use analog chemical signals. This makes it hard to replicate brains in computers because the current generation of computers is completely digital. However, the characteristics of these parts can be abstracted into mathematical functions that closely capture the essence of the neuron's operations.<br />
<br />
神经元及其突触的生物学过程极其复杂，难以人工模拟。生物大脑的一个关键特征是，神经元的所有处理过程都使用模拟化学信号。这使得很难在计算机中复制大脑，因为目前的计算机是完全数字化的。然而，这些部分的特征可以抽象为数学函数，这些函数紧密地抓住了神经元运行的本质。<br />
<br />
The goal of neuromorphic computing is not to perfectly mimic the brain and all of its functions, but instead to extract what is known of its structure and operations to be used in a practical computing system. No neuromorphic system will claim nor attempt to reproduce every element of neurons and synapses, but all adhere to the idea that computation is highly [[distributed processing|distributed]] throughout a series of small computing elements analogous to a neuron. While this sentiment is standard, researchers chase this goal with different methods.<ref>{{Cite journal | doi = 10.1088/1741-2560/13/5/051001| title = Large-scale neuromorphic computing systems| journal = Journal of Neural Engineering| volume = 13| pages = 1–15| year = 2016| last1 = Furber | first1 = Steve| issue = 5| pmid = 27529195| bibcode = 2016JNEng..13e1001F| doi-access = free}}</ref><br />
<br />
The goal of neuromorphic computing is not to perfectly mimic the brain and all of its functions, but instead to extract what is known of its structure and operations to be used in a practical computing system. No neuromorphic system will claim nor attempt to reproduce every element of neurons and synapses, but all adhere to the idea that computation is highly distributed throughout a series of small computing elements analogous to a neuron. While this sentiment is standard, researchers chase this goal with different methods.<br />
<br />
神经形态计算的目标不是完美地模拟大脑及其所有功能，而是提取已知的大脑结构和操作，用于实际的计算系统。没有哪个神经形态学系统会声称或试图复制神经元和突触的每一个元素，但所有人都坚持这样的观点，即计算是高度分布在一系列类似于神经元的小型计算元素中的。虽然这种情绪是标准的，但研究人员用不同的方法追求这一目标。<br />
<br />
==Examples==<br />
As early as 2006, researchers at [[Georgia Tech]] published a field programmable neural array.<ref>{{Cite book|title = A field programmable neural array|last1 = Farquhar|first1 = Ethan|date = May 2006|journal = IEEE International Symposium on Circuits and Systems|pages = 4114–4117|last2 = Hasler|first2 = Paul.|doi = 10.1109/ISCAS.2006.1693534|isbn = 978-0-7803-9389-9|s2cid = 206966013}}</ref> This chip was the first in a line of increasingly complex arrays of floating gate transistors that allowed programmability of charge on the gates of [[MOSFET]]s to model the channel-ion characteristics of neurons in the brain and was one of the first cases of a silicon programmable array of neurons.<br />
<br />
As early as 2006, researchers at Georgia Tech published a field programmable neural array. This chip was the first in a line of increasingly complex arrays of floating gate transistors that allowed programmability of charge on the gates of MOSFETs to model the channel-ion characteristics of neurons in the brain and was one of the first cases of a silicon programmable array of neurons.<br />
<br />
早在2006年，佐治亚理工学院的研究人员就发表了现场可编程神经阵列。这种芯片是一系列越来越复杂的浮栅晶体管阵列中的第一个，这些晶体管可以在 mosfet 的栅极上编程来模拟大脑中神经元的通道离子特性，这也是硅可编程神经元阵列的首批案例之一。<br />
<br />
In November 2011, a group of [[MIT]] researchers created a computer chip that mimics the analog, ion-based communication in a synapse between two neurons using 400 transistors and standard [[CMOS]] manufacturing techniques.<ref>{{cite web|title=MIT creates "brain chip"|url=http://www.extremetech.com/extreme/105067-mit-creates-brain-chip|access-date=4 December 2012}}</ref><ref name="Neuromorphic silicon paper">{{cite journal|title=Neuromorphic silicon neurons and large-scale neural networks: challenges and opportunities|doi=10.3389/fnins.2011.00108|pmid=21991244|pmc=3181466|volume=5|pages=108|journal=Frontiers in Neuroscience|year=2011|last1=Poon|first1=Chi-Sang|last2=Zhou|first2=Kuan|doi-access=free}}</ref><br />
<br />
In November 2011, a group of MIT researchers created a computer chip that mimics the analog, ion-based communication in a synapse between two neurons using 400 transistors and standard CMOS manufacturing techniques.<br />
<br />
2011年11月，麻省理工学院的一组研究人员发明了一种计算机芯片，该芯片使用400个晶体管和标准的 CMOS 制造技术，在两个神经元之间的突触中模拟模拟基于离子的通讯。<br />
<br />
In June 2012, [[spintronic]] researchers at [[Purdue University]] presented a paper on the design of a neuromorphic chip using [[Spin valve|lateral spin valve]]s and [[memristor]]s. They argue that the architecture works similarly to neurons and can therefore be used to test methods of reproducing the brain's processing. In addition, these chips are significantly more energy-efficient than conventional ones.<ref name="Spin Devices Prop">{{Cite arXiv|title=Proposal For Neuromorphic Hardware Using Spin Devices|eprint=1206.3227|last1=Sharad|first1=Mrigank|last2=Augustine|first2=Charles|last3=Panagopoulos|first3=Georgios|last4=Roy|first4=Kaushik|class=cond-mat.dis-nn|year=2012}}</ref><br />
<br />
In June 2012, spintronic researchers at Purdue University presented a paper on the design of a neuromorphic chip using lateral spin valves and memristors. They argue that the architecture works similarly to neurons and can therefore be used to test methods of reproducing the brain's processing. In addition, these chips are significantly more energy-efficient than conventional ones.<br />
<br />
2012年6月，普渡大学的自旋电子学研究人员发表了一篇关于利用侧向自旋阀和记忆电阻器设计神经形态芯片的论文。他们认为，这种结构的工作原理与神经元相似，因此可以用来测试复制大脑处理过程的方法。此外，这些芯片明显比传统芯片更节能。<br />
<br />
Research at [[HP Labs]] on Mott memristors has shown that while they can be non-[[Volatile memory|volatile]], the volatile behavior exhibited at temperatures significantly below the [[phase transition]] temperature can be exploited to fabricate a [[neuristor]],<ref name=":0" /> a biologically-inspired device that mimics behavior found in neurons.<ref name=":0">{{Cite journal | doi = 10.1038/nmat3510| pmid = 23241533| title = A scalable neuristor built with Mott memristors| journal = Nature Materials| volume = 12| issue = 2| pages = 114–7| year = 2012| last1 = Pickett | first1 = M. D. | last2 = Medeiros-Ribeiro | first2 = G. | last3 = Williams | first3 = R. S. | bibcode = 2013NatMa..12..114P| s2cid = 16271627| url = https://semanticscholar.org/paper/b6ba6f496ace2b947f111059663e76bb60e9efeb}}</ref> In September 2013, they presented models and simulations that show how the spiking behavior of these neuristors can be used to form the components required for a [[Turing machine]].<ref>{{cite journal|doi=10.1088/0957-4484/24/38/384002|title=Phase transitions enable computational universality in neuristor-based cellular automata|author1=Matthew D Pickett|author2=R Stanley Williams|name-list-style=amp|date=September 2013|publisher=IOP Publishing Ltd|journal=Nanotechnology|volume=24|issue=38|pmid=23999059|bibcode=2013Nanot..24L4002P|at=384002}}</ref><br />
<br />
Research at HP Labs on Mott memristors has shown that while they can be non-volatile, the volatile behavior exhibited at temperatures significantly below the phase transition temperature can be exploited to fabricate a neuristor, a biologically-inspired device that mimics behavior found in neurons. In September 2013, they presented models and simulations that show how the spiking behavior of these neuristors can be used to form the components required for a Turing machine.<br />
<br />
惠普实验室在莫特记忆电阻器上的研究表明，尽管它们可以是非挥发性的，但是在相变温度以下显示的挥发性行为可以被用来制造神经元电阻器，这是一种模仿神经元行为的生物设备。2013年9月，他们展示了模型和仿真，展示了这些神经元的尖峰行为是如何被用来形成图灵机所需的元件的。<br />
<br />
[[Neurogrid]], built by ''Brains in Silicon'' at [[Stanford University]],<ref>{{cite journal|last1=Boahen|first1=Kwabena|title=Neurogrid: A Mixed-Analog-Digital Multichip System for Large-Scale Neural Simulations|journal=Proceedings of the IEEE|date=24 April 2014|volume=102|issue=5|pages=699–716|doi=10.1109/JPROC.2014.2313565|s2cid=17176371}}</ref> is an example of hardware designed using neuromorphic engineering principles. The circuit board is composed of 16 custom-designed chips, referred to as NeuroCores. Each NeuroCore's analog circuitry is designed to emulate neural elements for 65536 neurons, maximizing energy efficiency. The emulated neurons are connected using digital circuitry designed to maximize spiking throughput.<ref>{{cite journal|doi=10.1038/503022a|pmid = 24201264|title = Neuroelectronics: Smart connections|journal = Nature|volume = 503|issue = 7474|pages = 22–4|year = 2013|last1 = Waldrop|first1 = M. Mitchell|bibcode = 2013Natur.503...22W|doi-access = free}}</ref><ref>{{cite journal|doi=10.1109/JPROC.2014.2313565|title = Neurogrid: A Mixed-Analog-Digital Multichip System for Large-Scale Neural Simulations|journal = Proceedings of the IEEE|volume = 102|issue = 5|pages = 699–716|year = 2014|last1 = Benjamin|first1 = Ben Varkey|last2 = Peiran Gao|last3 = McQuinn|first3 = Emmett|last4 = Choudhary|first4 = Swadesh|last5 = Chandrasekaran|first5 = Anand R.|last6 = Bussat|first6 = Jean-Marie|last7 = Alvarez-Icaza|first7 = Rodrigo|last8 = Arthur|first8 = John V.|last9 = Merolla|first9 = Paul A.|last10 = Boahen|first10 = Kwabena|s2cid = 17176371}}</ref><br />
<br />
Neurogrid, built by Brains in Silicon at Stanford University, is an example of hardware designed using neuromorphic engineering principles. The circuit board is composed of 16 custom-designed chips, referred to as NeuroCores. Each NeuroCore's analog circuitry is designed to emulate neural elements for 65536 neurons, maximizing energy efficiency. The emulated neurons are connected using digital circuitry designed to maximize spiking throughput.<br />
<br />
神经网格是由斯坦福大学硅谷大脑公司建立的，是一个使用神经形态工程原理设计硬件的例子。该电路板由16个定制设计的芯片组成，称为 NeuroCores。每个 NeuroCore 的模拟电路被设计为模拟65536个神经元的神经元元件，最大限度地提高能量效率。模拟的神经元通过设计的数字电路连接，以最大化脉冲吞吐量。<br />
<br />
A research project with implications for neuromorphic engineering is the [[Human Brain Project]] that is attempting to simulate a complete human brain in a supercomputer using biological data. It is made up of a group of researchers in neuroscience, medicine, and computing.<ref>{{cite web|title=Involved Organizations|url=http://www.humanbrainproject.eu/partners.html|access-date=22 February 2013|url-status=dead|archive-url=https://web.archive.org/web/20130302142627/http://www.humanbrainproject.eu/partners.html|archive-date=2 March 2013}}</ref> [[Henry Markram]], the project's co-director, has stated that the project proposes to establish a foundation to explore and understand the brain and its diseases, and to use that knowledge to build new computing technologies. The three primary goals of the project are to better understand how the pieces of the brain fit and work together, to understand how to objectively diagnose and treat brain diseases, and to use the understanding of the human brain to develop neuromorphic computers. That the simulation of a complete human brain will require a supercomputer a thousand times more powerful than today's encourages the current focus on neuromorphic computers.<ref>{{cite web|title=Human Brain Project|url=http://www.humanbrainproject.eu|access-date=22 February 2013}}</ref> $1.3 billion has been allocated to the project by The [[European Commission]].<ref>{{cite web|title=The Human Brain Project and Recruiting More Cyberwarriors|url=http://www.marketplace.org/topics/tech/human-brain-project-and-recruiting-more-cyberwarriors|access-date=22 February 2013|date=January 29, 2013}}</ref><br />
<br />
A research project with implications for neuromorphic engineering is the Human Brain Project that is attempting to simulate a complete human brain in a supercomputer using biological data. It is made up of a group of researchers in neuroscience, medicine, and computing. Henry Markram, the project's co-director, has stated that the project proposes to establish a foundation to explore and understand the brain and its diseases, and to use that knowledge to build new computing technologies. The three primary goals of the project are to better understand how the pieces of the brain fit and work together, to understand how to objectively diagnose and treat brain diseases, and to use the understanding of the human brain to develop neuromorphic computers. That the simulation of a complete human brain will require a supercomputer a thousand times more powerful than today's encourages the current focus on neuromorphic computers. $1.3 billion has been allocated to the project by The European Commission.<br />
<br />
一个对神经形态工程有影响的研究项目是人脑项目，它试图用生物数据在超级计算机中模拟完整的人脑。它由一群神经科学、医学和计算机科学的研究人员组成。该项目的联合主管亨利•马克拉姆(Henry Markram)表示，该项目提议建立一个基础，以探索和了解大脑及其疾病，并利用这些知识构建新的计算机技术。这个项目的三个主要目标是: 更好地理解大脑的各个部分是如何相互匹配和协同工作的; 理解如何客观地诊断和治疗脑部疾病; 以及利用对人类大脑的理解来开发神经形态计算机。模拟一个完整的人类大脑需要一台比现在强大一千倍的超级计算机，这鼓励了当前对神经形态计算机的关注。欧洲委员会已经拨款13亿美元用于这个项目。<br />
<br />
Other research with implications for neuromorphic engineering involves the [[BRAIN Initiative]]<ref name="economist">[https://www.economist.com/news/science-and-technology/21582495-computers-will-help-people-understand-brains-better-and-understanding-brains Neuromorphic computing: The machine of a new soul], The Economist, 2013-08-03</ref> and the [[TrueNorth]] chip from [[IBM]].<ref>{{cite journal|last1=Modha|first1=Dharmendra|title=A million spiking-neuron integrated circuit with a scalable communication network and interface|journal=Science|date=Aug 2014|volume=345|issue=6197|pages=668–673|doi=10.1126/science.1254642|pmid=25104385|bibcode=2014Sci...345..668M|s2cid=12706847}}</ref> Neuromorphic devices have also been demonstrated using nanocrystals, nanowires, and conducting polymers.<ref>{{Cite web|url=http://jessamynfairfield.com/wp-content/uploads/2017/03/PWMar17Fairfield.pdf|title=Smarter Machines|last=Fairfield|first=Jessamyn|date=March 1, 2017}}</ref><br />
<br />
Other research with implications for neuromorphic engineering involves the BRAIN InitiativeNeuromorphic computing: The machine of a new soul, The Economist, 2013-08-03 and the TrueNorth chip from IBM. Neuromorphic devices have also been demonstrated using nanocrystals, nanowires, and conducting polymers.<br />
<br />
其他与神经形态工程有关的研究还包括 BRAIN initiativity/euromorphic computing: The machine of a new soul，The Economist，2013-08-03 and The TrueNorth chip from IBM。神经形态学设备也已经被证明使用纳米晶体、纳米线和导电聚合物。<br />
<br />
[[Intel]] unveiled its neuromorphic research chip, called “[[Intel Loihi|Loihi]]”, in October 2017. The chip uses an asynchronous [[spiking neural network]] (SNN) to implement adaptive self-modifying event-driven fine-grained parallel computations used to implement learning and inference with high efficiency.<ref>{{cite journal |last1=Davies |first1=Mike |title=Loihi: A Neuromorphic Manycore Processor with On-Chip Learning |journal=IEEE Micro |date=January 16, 2018 |volume=38 |issue=1 |pages=82–99 |display-authors=etal|doi=10.1109/MM.2018.112130359 |s2cid=3608458 }}</ref><ref name="Morris2017">{{cite web |last1=Morris |first1=John |title=Why Intel built a neuromorphic chip |url=https://www.zdnet.com/article/why-intel-built-a-neuromorphic-chip/ |website=ZDNet |access-date=17 August 2018 |language=en}}</ref><br />
<br />
Intel unveiled its neuromorphic research chip, called “Loihi”, in October 2017. The chip uses an asynchronous spiking neural network (SNN) to implement adaptive self-modifying event-driven fine-grained parallel computations used to implement learning and inference with high efficiency.<br />
<br />
2017年10月，英特尔公布了其神经形态研究芯片“ Loihi”。该芯片采用异步脉冲神经网络(SNN)实现自适应自修改事件驱动的细粒度并行计算，实现了高效的学习和推理。<br />
<br />
[[IMEC]], a Belgium-based nanoelectronics research center, demonstrated the world's first self-learning neuromorphic chip. The brain-inspired chip, based on OxRAM technology, has the capability of self-learning and has been demonstrated to have the ability to compose music.<ref>{{cite web |title=Imec demonstrates self-learning neuromorphic chip that composes music |url=https://www.imec-int.com/en/articles/imec-demonstrates-self-learning-neuromorphic-chip-that-composes-music |website=IMEC International |access-date=1 October 2019}}</ref> IMEC released the 30-second tune composed by the prototype. The chip was sequentially loaded with songs in the same time signature and style. The songs were old Belgian and French flute minuets, from which the chip learned the rules at play and then applied them.<ref>{{cite web|last1=Bourzac|first1=Katherine|title=A Neuromorphic Chip That Makes Music|url=https://spectrum.ieee.org/a-neuromorphic-chip-that-makes-music|url-status=live|access-date=1 October 2019|website=IEEE Spectrum|date=May 23, 2017}}</ref><br />
<br />
IMEC, a Belgium-based nanoelectronics research center, demonstrated the world's first self-learning neuromorphic chip. The brain-inspired chip, based on OxRAM technology, has the capability of self-learning and has been demonstrated to have the ability to compose music. IMEC released the 30-second tune composed by the prototype. The chip was sequentially loaded with songs in the same time signature and style. The songs were old Belgian and French flute minuets, from which the chip learned the rules at play and then applied them.<br />
<br />
比利时的一个纳米电子学研究中心 IMEC 展示了世界上第一个自学习神经形态芯片。这种基于 OxRAM 技术的大脑启发芯片具有自学习能力，并已被证明具有创作音乐的能力。IMEC 发布了由原型机谱写的30秒曲调。芯片按顺序加载同时签名和风格的歌曲。歌曲是古老的比利时和法国长笛小步舞曲，筹码从中学习游戏规则，然后应用它们。<br />
<br />
[[Blue Brain Project|The Blue Brain Project]], led by Henry Markram, aims to build biologically detailed digital reconstructions and simulations of the mouse brain. The Blue Brain Project has created in silico models of rodent brains, while attempting to replicate as many details about its biology as possible. The supercomputer-based simulations offer new perspectives on understanding the structure and functions of the brain.<br />
<br />
The Blue Brain Project, led by Henry Markram, aims to build biologically detailed digital reconstructions and simulations of the mouse brain. The Blue Brain Project has created in silico models of rodent brains, while attempting to replicate as many details about its biology as possible. The supercomputer-based simulations offer new perspectives on understanding the structure and functions of the brain.<br />
<br />
由亨利 · 马克拉姆领导的蓝脑计划旨在建立小鼠大脑的生物细节数字重建和模拟。蓝色大脑项目已经在啮齿动物大脑的电子模型中创建，同时试图尽可能多地复制其生物学细节。基于超级计算机的模拟为理解大脑的结构和功能提供了新的视角。<br />
<br />
The European Union funded a series of projects at the University of Heidelberg, which led to the development of <br />
[[BrainScaleS]] (brain-inspired multiscale computation in neuromorphic hybrid systems), a hybrid analog [[neuromorphic]] supercomputer located at Heidelberg University, Germany. It was developed as part of the [[Human Brain Project]] neuromorphic computing platform and is the complement to the [[SpiNNaker]] supercomputer (which is based on digital technology). The architecture used in BrainScaleS mimics biological neurons and their connections on a physical level; additionally, since the components are made of silicon, these model neurons operate on average 864 times (24 hours of real time is 100 seconds in the machine simulation) that of their biological counterparts.<ref>{{Cite web|date=2016-03-21|title=Beyond von Neumann, Neuromorphic Computing Steadily Advances|url=https://www.hpcwire.com/2016/03/21/lacking-breakthrough-neuromorphic-computing-steadily-advance/|access-date=2021-10-08|website=HPCwire|language=en-US}}</ref><br />
<br />
The European Union funded a series of projects at the University of Heidelberg, which led to the development of <br />
BrainScaleS (brain-inspired multiscale computation in neuromorphic hybrid systems), a hybrid analog neuromorphic supercomputer located at Heidelberg University, Germany. It was developed as part of the Human Brain Project neuromorphic computing platform and is the complement to the SpiNNaker supercomputer (which is based on digital technology). The architecture used in BrainScaleS mimics biological neurons and their connections on a physical level; additionally, since the components are made of silicon, these model neurons operate on average 864 times (24 hours of real time is 100 seconds in the machine simulation) that of their biological counterparts.<br />
<br />
欧盟资助了海德堡大学的一系列项目，这些项目导致了 BrainScaleS (神经形态混合系统中受大脑启发的多尺度计算)的发展，这是一台位于德国海德堡大学的混合模拟神经形态超级计算机。它是作为人脑计划神经形态计算平台的一部分而开发的，是 SpiNNaker 超级计算机(基于数字技术)的补充。大脑尺度中使用的体系结构模拟了生物神经元及其在物理层面上的连接; 此外，由于这些组件是由硅制成的，这些模型神经元平均运行864倍(在机器模拟中，24小时的实时时间是100秒)。<br />
<br />
===Neuromorphic sensors===<br />
The concept of neuromorphic systems can be extended to sensors (not just to computation). An example of this applied to detecting [[light]] is the [[retinomorphic sensor]] or, when employed in an array, the [[event camera]].<br />
<br />
The concept of neuromorphic systems can be extended to sensors (not just to computation). An example of this applied to detecting light is the retinomorphic sensor or, when employed in an array, the event camera.<br />
<br />
= = = 神经形态传感器 = = = 神经形态系统的概念可以扩展到传感器(而不仅仅是计算)。用于检测光线的一个例子是视网膜变形传感器，或者在阵列中使用的事件摄像机。<br />
<br />
==Ethical considerations==<br />
While the interdisciplinary concept of neuromorphic engineering is relatively new, many of the same ethical considerations apply to neuromorphic systems as apply to [[human-like machines]] and [[artificial intelligence]] in general. However, the fact that neuromorphic systems are designed to mimic a [[human brain]] gives rise to unique ethical questions surrounding their usage.<br />
<br />
While the interdisciplinary concept of neuromorphic engineering is relatively new, many of the same ethical considerations apply to neuromorphic systems as apply to human-like machines and artificial intelligence in general. However, the fact that neuromorphic systems are designed to mimic a human brain gives rise to unique ethical questions surrounding their usage.<br />
<br />
虽然神经形态工程的跨学科概念相对较新，但许多同样的伦理考虑适用于神经形态系统，就像适用于类人机器和一般人工智能一样。然而，神经形态系统是为了模仿人类大脑而设计的这一事实引起了围绕其使用的独特的伦理问题。<br />
<br />
However, the practical debate is that neuromorphic hardware as well as artificial "neural networks" are immensely simplified models of how the brain operates or processes information at a much lower [[complex system|complexity in terms of size and functional technology]] and a much more regular structure in terms of [[brain connectivity|connectivity]]. Comparing [[neuromorphic chip]]s to the brain is a very crude comparison similar to comparing a plane to a bird just because they both have wings and a tail. The fact is that neural cognitive systems are many orders of magnitude more [[Energy efficiency (physics)|energy-]] and compute-efficient than current state-of-the-art AI and neuromorphic engineering is an attempt to narrow this gap by inspiring from the brain's mechanism just like many engineering designs have [[bioengineering|bio-inspired features]].<br />
<br />
However, the practical debate is that neuromorphic hardware as well as artificial "neural networks" are immensely simplified models of how the brain operates or processes information at a much lower complexity in terms of size and functional technology and a much more regular structure in terms of connectivity. Comparing neuromorphic chips to the brain is a very crude comparison similar to comparing a plane to a bird just because they both have wings and a tail. The fact is that neural cognitive systems are many orders of magnitude more energy- and compute-efficient than current state-of-the-art AI and neuromorphic engineering is an attempt to narrow this gap by inspiring from the brain's mechanism just like many engineering designs have bio-inspired features.<br />
<br />
然而，实际上的争论是，神经形态硬件和人工“神经网络”是大脑如何运作或处理信息的极其简化的模型，在大小和功能技术方面的复杂性要低得多，在连接方面的结构则更加规则。将神经形态芯片与大脑进行比较是一种非常粗糙的比较，类似于仅仅因为一架飞机有翅膀和一条尾巴就将它与一只鸟进行比较。事实上，神经认知系统比当前最先进的人工智能具有更多的能量和计算效率，而神经形态工程是一种通过从大脑机制中激发灵感来缩小这种差距的尝试，就像许多工程设计具有生物启发的特征一样。数量级。<br />
<br />
===Democratic concerns===<br />
Significant ethical limitations may be placed on neuromorphic engineering due to public perception.<ref>{{Cite report|url=https://ai100.stanford.edu/sites/g/files/sbiybj9861/f/ai_100_report_0831fnl.pdf|title=Artificial Intelligence and Life in 2030|author=2015 Study Panel|date=September 2016|work=One Hundred Year Study on Artificial Intelligence (AI100)|publisher=Stanford University}}</ref> Special [[Eurobarometer]] 382: Public Attitudes Towards Robots, a survey conducted by the European Commission, found that 60% of [[European Union]] citizens wanted a ban of robots in the care of children, the elderly, or the disabled. Furthermore, 34% were in favor of a ban on robots in education, 27% in healthcare, and 20% in leisure. The European Commission classifies these areas as notably “human.” The report cites increased public concern with robots that are able to mimic or replicate human functions. Neuromorphic engineering, by definition, is designed to replicate the function of the human brain.<ref name=":1">{{Cite web|url=http://ec.europa.eu/commfrontoffice/publicopinion/archives/ebs/ebs_382_en.pdf|title=Special Eurobarometer 382: Public Attitudes Towards Robots|last=European Commission|date=September 2012|website=European Commission}}</ref><br />
<br />
Significant ethical limitations may be placed on neuromorphic engineering due to public perception. Special Eurobarometer 382: Public Attitudes Towards Robots, a survey conducted by the European Commission, found that 60% of European Union citizens wanted a ban of robots in the care of children, the elderly, or the disabled. Furthermore, 34% were in favor of a ban on robots in education, 27% in healthcare, and 20% in leisure. The European Commission classifies these areas as notably “human.” The report cites increased public concern with robots that are able to mimic or replicate human functions. Neuromorphic engineering, by definition, is designed to replicate the function of the human brain.<br />
<br />
由于公众的认知，神经形态工程学可能会受到严重的伦理限制。欧盟委员会进行的一项调查发现，60% 的欧盟公民希望禁止照顾儿童、老人或残疾人的机器人。此外，34% 的人支持禁止机器人用于教育，27% 的人支持医疗保健，20% 的人支持休闲。欧盟委员会将这些地区明显归类为“人类”报告指出，公众越来越关注能够模仿或复制人类功能的机器人。神经形态工程，顾名思义，是为了复制人脑的功能而设计的。<br />
<br />
The democratic concerns surrounding neuromorphic engineering are likely to become even more profound in the future. The European Commission found that EU citizens between the ages of 15 and 24 are more likely to think of robots as human-like (as opposed to instrument-like) than EU citizens over the age of 55. When presented an image of a robot that had been defined as human-like, 75% of EU citizens aged 15–24 said it corresponded with the idea they had of robots while only 57% of EU citizens over the age of 55 responded the same way. The human-like nature of neuromorphic systems, therefore, could place them in the categories of robots many EU citizens would like to see banned in the future.<ref name=":1" /><br />
<br />
The democratic concerns surrounding neuromorphic engineering are likely to become even more profound in the future. The European Commission found that EU citizens between the ages of 15 and 24 are more likely to think of robots as human-like (as opposed to instrument-like) than EU citizens over the age of 55. When presented an image of a robot that had been defined as human-like, 75% of EU citizens aged 15–24 said it corresponded with the idea they had of robots while only 57% of EU citizens over the age of 55 responded the same way. The human-like nature of neuromorphic systems, therefore, could place them in the categories of robots many EU citizens would like to see banned in the future.<br />
<br />
围绕神经形态工程的民主关注可能在未来变得更加深刻。欧盟委员会(European Commission)发现，15至24岁的欧盟公民比55岁以上的欧盟公民更有可能认为机器人像人(而不是像仪器)。当看到一张被定义为“类人”的机器人图片时，年龄在15岁至24岁之间的欧盟公民中有75% 的人表示，这与他们对机器人的想法相符，而55岁以上的欧盟公民中只有57% 的人有同样的反应。因此，类似人类的神经形态系统，可以把它们归入许多欧盟公民希望在未来禁止使用的机器人类别。<br />
<br />
===Personhood===<br />
As neuromorphic systems have become increasingly advanced, some scholars{{who|date=August 2021}} have advocated for granting [[personhood]] rights to these systems. If the brain is what grants humans their personhood, to what extent does a neuromorphic system have to mimic the human brain to be granted personhood rights? Critics of technology development in the [[Human Brain Project]], which aims to advance brain-inspired computing, have argued that advancement in neuromorphic computing could lead to machine consciousness or personhood.<ref>{{Cite journal|last=Aicardi|first=Christine|date=September 2018|title=Accompanying technology development in the Human Brain Project: From foresight to ethics management|journal=Futures|volume=102|pages=114–124|doi=10.1016/j.futures.2018.01.005|doi-access=free}}</ref> If these systems are to be treated as people, critics argue, then many tasks humans perform using neuromorphic systems, including the act of termination of neuromorphic systems, may be morally impermissible as these acts would violate the autonomy of the neuromorphic systems.<ref>{{Cite journal|last=Lim|first=Daniel|date=2014-06-01|title=Brain simulation and personhood: a concern with the Human Brain Project|journal=Ethics and Information Technology|language=en|volume=16|issue=2|pages=77–89|doi=10.1007/s10676-013-9330-5|s2cid=17415814|issn=1572-8439}}</ref><br />
<br />
As neuromorphic systems have become increasingly advanced, some scholars have advocated for granting personhood rights to these systems. If the brain is what grants humans their personhood, to what extent does a neuromorphic system have to mimic the human brain to be granted personhood rights? Critics of technology development in the Human Brain Project, which aims to advance brain-inspired computing, have argued that advancement in neuromorphic computing could lead to machine consciousness or personhood. If these systems are to be treated as people, critics argue, then many tasks humans perform using neuromorphic systems, including the act of termination of neuromorphic systems, may be morally impermissible as these acts would violate the autonomy of the neuromorphic systems.<br />
<br />
= = = 人格 = = 随着神经形态系统的日益发展，一些学者主张赋予这些系统人格权。如果是大脑赋予了人类人格，那么神经形态系统在多大程度上必须模仿人类大脑才能被赋予人格权利？“人脑计划”旨在推进以大脑为灵感的计算机技术发展，该计划的批评者认为，神经形态计算机技术的进步可能导致机器意识或人格的形成。批评家认为，如果这些系统被当作人来对待，那么人类使用神经形态系统执行的许多任务，包括终止神经形态系统的行为，可能在道德上是不允许的，因为这些行为将违反神经形态系统的自主性。<br />
<br />
==Dual use (military applications)==<br />
The [[Joint Artificial Intelligence Center]], a branch of the U.S. military, is a center dedicated to the procurement and implementation of AI software and neuromorphic hardware for combat use. Specific applications include smart headsets/goggles and robots. JAIC intends to rely heavily on neuromorphic technology to connect "every fighter every shooter" within a network of neuromorphic-enabled units.<br />
<br />
The Joint Artificial Intelligence Center, a branch of the U.S. military, is a center dedicated to the procurement and implementation of AI software and neuromorphic hardware for combat use. Specific applications include smart headsets/goggles and robots. JAIC intends to rely heavily on neuromorphic technology to connect "every fighter every shooter" within a network of neuromorphic-enabled units.<br />
<br />
军民两用联合人工智能中心是美国军队的一个分支，专门从事采购和实施用于战斗的人工智能软件和神经形态硬件。具体应用包括智能耳机/护目镜和机器人。JAIC 打算严重依赖神经形态技术来连接神经形态单位网络中的“每个战士每个射手”。<br />
<br />
==Legal considerations==<br />
Skeptics have argued that there is no way to apply the electronic personhood, the concept of personhood that would apply to neuromorphic technology, legally. In a letter signed by 285 experts in law, robotics, medicine, and ethics opposing a European Commission proposal to recognize “smart robots” as legal persons, the authors write, “A legal status for a robot can’t derive from the [[Natural person|Natural Person]] model, since the robot would then hold [[human rights]], such as the right to dignity, the right to its integrity, the right to remuneration or the right to citizenship, thus directly confronting the Human rights. This would be in contradiction with the [[Charter of Fundamental Rights of the European Union]] and the [[Convention for the Protection of Human Rights and Fundamental Freedoms]].”<ref>{{Cite web|url=http://www.robotics-openletter.eu/|title=Robotics Openletter {{!}} Open letter to the European Commission|language=fr-FR|access-date=2019-05-10}}</ref><br />
<br />
Skeptics have argued that there is no way to apply the electronic personhood, the concept of personhood that would apply to neuromorphic technology, legally. In a letter signed by 285 experts in law, robotics, medicine, and ethics opposing a European Commission proposal to recognize “smart robots” as legal persons, the authors write, “A legal status for a robot can’t derive from the Natural Person model, since the robot would then hold human rights, such as the right to dignity, the right to its integrity, the right to remuneration or the right to citizenship, thus directly confronting the Human rights. This would be in contradiction with the Charter of Fundamental Rights of the European Union and the Convention for the Protection of Human Rights and Fundamental Freedoms.”<br />
<br />
= = 法律方面的考虑 = = 怀疑论者认为，在法律上没有办法应用电子人格，这个人格概念将适用于神经形态技术。在一封由285名法律、机器人、医学和伦理学专家签名的信中，作者们反对欧盟委员会承认“智能机器人”为法人的提议，他们写道，“机器人的法律地位不能从自然人模型中推导出来，因为机器人将拥有人权，如尊严权、完整权、报酬权或公民权，从而直接面对人权。这将有悖于《欧洲联盟基本权利宪章和《保护人权和基本自由公约》。”<br />
<br />
=== Ownership and property rights===<br />
There is significant legal debate around property rights and artificial intelligence. In ''Acohs Pty Ltd v. Ucorp Pty Ltd'', Justice Christopher Jessup of the [[Federal Court of Australia]] found that the [[source code]] for [[Material safety data sheets|Material Safety Data Sheets]] could not be [[Copyright law of Australia|copyrighted]] as it was generated by a [[software interface]] rather than a human author.<ref>{{Cite web|url=http://www.lavan.com.au/advice/intellectual_property_technology/copyright_in_source_code_and_digital_products|title=Copyright in source code and digital products|last=Lavan|website=Lavan|language=en|access-date=2019-05-10}}</ref> The same question may apply to neuromorphic systems: if a neuromorphic system successfully mimics a human brain and produces a piece of original work, who, if anyone, should be able to claim ownership of the work?<ref>{{cite journal |last1=Eshraghian|first1=Jason K. |title=Human Ownership of Artificial Creativity |journal=Nature Machine Intelligence |date=9 March 2020 |volume=2 |pages=157–160 |doi=10.1038/s42256-020-0161-x}}</ref><br />
<br />
There is significant legal debate around property rights and artificial intelligence. In Acohs Pty Ltd v. Ucorp Pty Ltd, Justice Christopher Jessup of the Federal Court of Australia found that the source code for Material Safety Data Sheets could not be copyrighted as it was generated by a software interface rather than a human author. The same question may apply to neuromorphic systems: if a neuromorphic system successfully mimics a human brain and produces a piece of original work, who, if anyone, should be able to claim ownership of the work?<br />
<br />
= = = 所有权和财产权 = = = 围绕财产权和人工智能有着重大的法律争论。在 Acohs Pty Ltd 诉 Ucorp Pty Ltd 一案中，澳大利亚联邦法院的克里斯托弗 · 杰瑟普法官发现，材料安全数据表的源代码不能受版权保护，因为它是由软件界面而不是人工作者生成的。同样的问题可能也适用于神经形态系统: 如果一个神经形态系统成功地模仿了人类的大脑并产生了一部原创作品，那么谁，如果有人，应该声称拥有这部作品的所有权？<br />
<br />
==Neuromemristive systems==<br />
Neuromemristive systems are a subclass of neuromorphic computing systems that focus on the use of [[memristors]] to implement [[neuroplasticity]]. While neuromorphic engineering focuses on mimicking biological behavior, neuromemristive systems focus on abstraction.<ref>{{Cite web|url=https://digitalops.sandia.gov/Mediasite/Play/a10cf6ceb55d47608bb8326dd00e46611d|title=002.08 N.I.C.E. Workshop 2014: Towards Intelligent Computing with Neuromemristive Circuits and Systems - Feb. 2014|website=digitalops.sandia.gov|access-date=2019-08-26}}</ref> For example, a neuromemristive system may replace the details of a [[Cerebral cortex|cortical]] microcircuit's behavior with an abstract neural network model.<ref>C. Merkel and D. Kudithipudi, "Neuromemristive extreme learning machines for pattern classification," ISVLSI, 2014.</ref><br />
<br />
Neuromemristive systems are a subclass of neuromorphic computing systems that focus on the use of memristors to implement neuroplasticity. While neuromorphic engineering focuses on mimicking biological behavior, neuromemristive systems focus on abstraction. For example, a neuromemristive system may replace the details of a cortical microcircuit's behavior with an abstract neural network model.C. Merkel and D. Kudithipudi, "Neuromemristive extreme learning machines for pattern classification," ISVLSI, 2014.<br />
<br />
神经记忆电阻系统是神经形态计算系统的一个亚类，主要研究利用记忆电阻器实现神经可塑性。神经形态工程的重点是模拟生物行为，而神经记忆电阻系统的重点是提取。例如，一个神经记忆系统可以用一个抽象的神经网络模型取代皮层微电路的行为细节。默克尔和 d. Kudithipudi，“用于模式分类的神经记忆极端学习机器”，ISVLSI，2014。<br />
<br />
There exist several neuron inspired threshold logic functions<ref name="Maan 1–13" /> implemented with memristors that have applications in high level [[pattern recognition]] applications. Some of the applications reported recently include [[speech recognition]],<ref>{{Cite journal|title = Memristor pattern recogniser: isolated speech word recognition|journal = Electronics Letters|pages = 1370–1372|volume = 51|issue = 17|doi = 10.1049/el.2015.1428|first1 = A.K.|last1 = Maan|first2 = A.P.|last2 = James|first3 = S.|last3 = Dimitrijev|year = 2015|bibcode = 2015ElL....51.1370M|hdl = 10072/140989|s2cid = 61454815|url = https://semanticscholar.org/paper/48d3ab11ec6e213b62f11eedcfb7b7febb058674|hdl-access = free}}</ref> [[face recognition]]<ref>{{Cite journal|title = Memristive Threshold Logic Face Recognition|journal = Procedia Computer Science|date = 2014-01-01|pages = 98–103|volume = 41|series = 5th Annual International Conference on Biologically Inspired Cognitive Architectures, 2014 BICA|doi = 10.1016/j.procs.2014.11.090|first1 = Akshay Kumar|last1 = Maan|first2 = Dinesh S.|last2 = Kumar|first3 = Alex Pappachen|last3 = James|doi-access = free}}</ref> and [[object recognition]].<ref>{{Cite journal|title = Memristive Threshold Logic Circuit Design of Fast Moving Object Detection|journal = IEEE Transactions on Very Large Scale Integration (VLSI) Systems|date = 2015-10-01|issn = 1063-8210|pages = 2337–2341|volume = 23|issue = 10|doi = 10.1109/TVLSI.2014.2359801|first1 = A.K.|last1 = Maan|first2 = D.S.|last2 = Kumar|first3 = S.|last3 = Sugathan|first4 = A.P.|last4 = James|arxiv = 1410.1267|s2cid = 9647290}}</ref> They also find applications in replacing conventional digital logic gates.<ref>{{Cite journal|title = Resistive Threshold Logic|journal = IEEE Transactions on Very Large Scale Integration (VLSI) Systems|date = 2014-01-01|issn = 1063-8210|pages = 190–195|volume = 22|issue = 1|doi = 10.1109/TVLSI.2012.2232946|first1 = A.P.|last1 = James|first2 = L.R.V.J.|last2 = Francis|first3 = D.S.|last3 = Kumar|arxiv = 1308.0090|s2cid = 7357110}}</ref><ref>{{Cite journal|title = Threshold Logic Computing: Memristive-CMOS Circuits for Fast Fourier Transform and Vedic Multiplication|journal = IEEE Transactions on Very Large Scale Integration (VLSI) Systems|date = 2015-11-01|issn = 1063-8210|pages = 2690–2694|volume = 23|issue = 11|doi = 10.1109/TVLSI.2014.2371857|first1 = A.P.|last1 = James|first2 = D.S.|last2 = Kumar|first3 = A.|last3 = Ajayan|arxiv = 1411.5255|s2cid = 6076956}}</ref><br />
<br />
There exist several neuron inspired threshold logic functions implemented with memristors that have applications in high level pattern recognition applications. Some of the applications reported recently include speech recognition, face recognition and object recognition. They also find applications in replacing conventional digital logic gates.<br />
<br />
利用记忆电阻器实现的神经元启发阈值逻辑函数在高级模式识别中有着广泛的应用。最近报道的一些应用包括语音识别、人脸识别和物体识别。它们还可以用来取代传统的数字逻辑门。<br />
<br />
For ideal passive memristive circuits there is an exact equation (Caravelli-Traversa-[[Di Ventra]] equation) for the internal memory of the circuit:<ref>{{cite journal |last=Caravelli |display-authors=etal|arxiv=1608.08651 |title=The complex dynamics of memristive circuits: analytical results and universal slow relaxation |year=2017 |doi=10.1103/PhysRevE.95.022140 |pmid= 28297937 |volume=95 |issue= 2 |pages= 022140 |journal=Physical Review E|bibcode=2017PhRvE..95b2140C |s2cid=6758362}}</ref><br />
<br />
For ideal passive memristive circuits there is an exact equation (Caravelli-Traversa-Di Ventra equation) for the internal memory of the circuit:<br />
<br />
对于理想的无源记忆电路，电路的内部记忆有一个精确的方程(Caravelli-Traversa-Di Ventra 方程) :<br />
<br />
:<math> \frac{d}{dt} \vec{W} = \alpha \vec{W}-\frac{1}{\beta} (I+\xi \Omega W)^{-1} \Omega \vec S </math><br />
<br />
:\frac{d}{dt} \vec{W} = \alpha \vec{W}-\frac{1}{\beta} (I+\xi \Omega W)^{-1} \Omega \vec S<br />
<br />
: \frac{d}{dt} \vec{W} = \alpha \vec{W}-\frac{1}{\beta} (I+\xi \Omega W)^{-1} \Omega \vec S<br />
<br />
as a function of the properties of the physical memristive network and the external sources. In the equation above, <math>\alpha</math> is the "forgetting" time scale constant, <math> \xi=r-1</math> and <math>r=\frac{R_\text{off}}{R_\text{on}}</math> is the ratio of ''off'' and ''on'' values of the limit resistances of the memristors, <math> \vec S </math> is the vector of the sources of the circuit and <math>\Omega</math> is a projector on the fundamental loops of the circuit. The constant <math>\beta</math> has the dimension of a voltage and is associated to the properties of the [[memristor]]; its physical origin is the charge mobility in the conductor. The diagonal matrix and vector <math>W=\operatorname{diag}(\vec W)</math> and <math>\vec W</math> respectively, are instead the internal value of the memristors, with values between 0 and 1. This equation thus requires adding extra constraints on the memory values in order to be reliable.<br />
<br />
<nowiki>as a function of the properties of the physical memristive network and the external sources. In the equation above, \alpha is the "forgetting" time scale constant, \xi=r-1 and r=\frac{R_\text{off}}{R_\text{on}} is the ratio of off and on values of the limit resistances of the memristors, \vec S is the vector of the sources of the circuit and \Omega is a projector on the fundamental loops of the circuit. The constant \beta has the dimension of a voltage and is associated to the properties of the memristor; its physical origin is the charge mobility in the conductor. The diagonal matrix and vector W=\operatorname{diag}(\vec W) and \vec W respectively, are instead the internal value of the memristors, with values between 0 and 1. This equation thus requires adding extra constraints on the memory values in order to be reliable.</nowiki><br />
<br />
<nowiki>作为物理记忆网络和外部源的性质的函数。在上述方程中，α 是“遗忘”时间尺度常数，xi = r-1，r = frac { r _ text { off }{ on }{ r _ text { on }}是记忆电阻器极限电阻的开关和开关值之比，vec s 是电路源的矢量，Omega 是电路基本环路的投影仪。常数 β 具有电压的尺寸，与记忆电阻器的特性有关; 它的物理起源是导体中的电荷迁移率。对角矩阵和向量 w = 操作者名{ diag }(vec w)和 vec w 分别是记忆电阻器的内值，值在0到1之间。因此，这个等式需要在内存值上添加额外的约束，以保证可靠性。</nowiki><br />
<br />
==See also ==<br />
{{Columns-list|colwidth=18em|<br />
* [[AI accelerator (computer hardware)|AI accelerator]]<br />
* [[Artificial brain]]<br />
* [[Biomorphic]]<br />
* [[Cognitive computer]]<br />
* [[Computation and Neural Systems]]<br />
* [[Differentiable programming]]<br />
* [[Event camera]]<br />
* [[Neurorobotics]]<br />
* [[Optical flow sensor]]<br />
* [[Physical neural network]]<br />
* [[SpiNNaker]]<br />
* [[SyNAPSE]]<br />
* [[Retinomorphic sensor]]<br />
* [[Vision chip]]<br />
* [[Vision processing unit]]<br />
* [[Zeroth (software)]]<br />
* [[Hardware for artificial intelligence]]<br />
}}<br />
<br />
{{Portal bar|Electronics}}<br />
<br />
==References==<br />
{{Reflist|40em}}<br />
<br />
==External links==<br />
<!--======================== {{No more links}} ============================<br />
| PLEASE BE CAUTIOUS IN ADDING MORE LINKS TO THIS ARTICLE. Wikipedia |<br />
| is not a collection of links nor should it be used for advertising. |<br />
| |<br />
| Excessive or inappropriate links WILL BE DELETED. |<br />
| See [[Wikipedia:External links]] & [[Wikipedia:Spam]] for details. |<br />
| |<br />
| If there are already plentiful links, please propose additions or |<br />
| replacements on this article's discussion page, or submit your link |<br />
| to the relevant category at the Open Directory Project (dmoz.org) |<br />
| and link back to that category using the {{dmoz}} template. |<br />
======================={{No more links}}=============================--><br />
{{Commons category|Neuromorphic Engineering}}<br />
*[https://web.archive.org/web/20150727034331/http://ine-web.org/workshops/workshops-overview Telluride Neuromorphic Engineering Workshop]<br />
*[https://archive.today/20130115190057/http://capocaccia.ethz.ch/ CapoCaccia Cognitive Neuromorphic Engineering Workshop]<br />
*[https://web.archive.org/web/20190716132350/http://www.ine-web.org/ Institute of Neuromorphic Engineering]<br />
*[http://www.ine-news.org/ INE news site].<br />
*[http://www.frontiersin.org/neuromorphic_engineering Frontiers in Neuromorphic Engineering Journal]<br />
*[http://www.cns.caltech.edu/ Computation and Neural Systems] department at the [[California Institute of Technology]].<br />
*[http://www.humanbrainproject.eu/ Human Brain Project official site]<br />
*[https://www.the-scientist.com/features/building-a-silicon-brain-65738 Building a Silicon Brain:] Computer chips based on biological neurons may help simulate larger and more-complex brain models. May 1, 2019. SANDEEP RAVINDRAN<br />
<br />
<br />
* Telluride Neuromorphic Engineering Workshop<br />
*CapoCaccia Cognitive Neuromorphic Engineering Workshop<br />
*Institute of Neuromorphic Engineering<br />
*INE news site.<br />
*Frontiers in Neuromorphic Engineering Journal<br />
*Computation and Neural Systems department at the California Institute of Technology.<br />
*Human Brain Project official site<br />
*Building a Silicon Brain: Computer chips based on biological neurons may help simulate larger and more-complex brain models. May 1, 2019. SANDEEP RAVINDRAN<br />
<br />
=<nowiki>= = 外部链接 =</nowiki>= <br />
*碲化物神经形态工程工作室<br />
*CapoCaccia 认知神经形态工程工作室<br />
*神经形态工程研究所<br />
*INE 新闻站点。<br />
*《神经形态工程学前沿》<br />
*加州理工学院计算与神经系统系。<br />
*人脑项目官方网站<br />
*打造硅脑: 基于生物神经元的计算机芯片可能有助于模拟更大、更复杂的大脑模型。2019年5月1日。SANDEEP RAVINDRAN<br />
<br />
{{Differentiable computing}}<br />
{{Authority control}}<br />
<br />
[[Category:Electrical engineering]]<br />
[[Category:Neuroscience]]<br />
[[分类:AI accelerators|*]]<br />
[[Category:Artificial intelligence]]<br />
[[Category:Robotics]]<br />
<br />
Category:Electrical engineering<br />
Category:Neuroscience<br />
*<br />
Category:Artificial intelligence<br />
Category:Robotics<br />
<br />
类别: 电气工程类别: 神经科学 * 类别: 人工智能类别: 机器人<br />
<br />
<noinclude><br />
<br />
<small>This page was moved from [[wikipedia:en:Neuromorphic engineering]]. Its edit history can be viewed at [[神经形态计算/edithistory]]</small></noinclude><br />
<br />
[[Category:待整理页面]]<br />
<br />
<references /></div>Shenky20https://wiki.swarma.org/index.php?title=%E7%A5%9E%E7%BB%8F%E5%BD%A2%E6%80%81%E8%AE%A1%E7%AE%97&diff=30173神经形态计算2022-04-13T07:34:59Z<p>Shenky20：</p>
<hr />
<div><br />
此词条由神经动力学模型读书会词条梳理志愿者Shenky20翻译审校，未经专家审校，带来阅读不便，请见谅。<br />
<br />
{{Use American English|date = January 2019}}<br />
{{Use mdy dates|date = January 2019}}<br />
<br />
'''Neuromorphic engineering''', also known as '''neuromorphic computing''',<ref>{{Cite journal | doi = 10.1145/2601069| title = Neuromorphic computing gets ready for the (really) big time| journal = [[Communications of the ACM]]| volume = 57| issue = 6| year = 2014| last1 = Monroe | first1 = D. | pages = 13–15| s2cid = 20051102}}</ref><ref>{{Cite journal | doi = 10.1088/0957-4484/21/17/175202| pmid = 20368686| title = Nanotube devices based crossbar architecture: Toward neuromorphic computing| journal = Nanotechnology| volume = 21| issue = 17| pages = 175202| year = 2010| last1 = Zhao | first1 = W. S. | last2 = Agnus | first2 = G. | last3 = Derycke | first3 = V. | last4 = Filoramo | first4 = A. | last5 = Bourgoin | first5 = J. -P. | last6 = Gamrat | first6 = C. | bibcode = 2010Nanot..21q5202Z| url = https://zenodo.org/record/3428659}}</ref><ref name="humanbrainproject">{{YouTube|id=6RoiZ90mGfw|title=The Human Brain Project SP 9: Neuromorphic Computing Platform}}</ref> is the use of [[very-large-scale integration]] (VLSI) systems containing electronic [[analog circuit]]s to mimic neuro-biological architectures present in the nervous system. A neuromorphic computer/chip is any device that uses physical artificial neurons (made from silicon) to do computations.<ref>{{cite journal|last1=Mead|first1=Carver|title=Neuromorphic electronic systems|journal=Proceedings of the IEEE|date=1990|volume=78|issue=10|pages=1629–1636|doi=10.1109/5.58356|url=https://authors.library.caltech.edu/53090/1/00058356.pdf}}</ref><ref name=":2" /> In recent times, the term ''neuromorphic'' has been used to describe analog, digital, mixed-mode analog/digital VLSI, and software systems that implement models of [[neural system]]s (for [[perception]], [[motor control]], or [[multisensory integration]]). The implementation of neuromorphic computing on the hardware level can be realized by oxide-based [[memristor]]s,<ref name="Maan 1–13">{{Cite journal|last1=Maan|first1=A. K.|last2=Jayadevi|first2=D. A.|last3=James|first3=A. P.|date=2016-01-01|title=A Survey of Memristive Threshold Logic Circuits|journal=IEEE Transactions on Neural Networks and Learning Systems|volume=PP|issue=99|pages=1734–1746|doi=10.1109/TNNLS.2016.2547842|pmid=27164608|issn=2162-237X|arxiv=1604.07121|bibcode=2016arXiv160407121M|s2cid=1798273}}</ref> spintronic memories, threshold switches, and [[transistor]]s.<ref>{{Cite journal|title = Mott Memory and Neuromorphic Devices|journal = Proceedings of the IEEE|date = 2015-08-01|issn = 0018-9219|pages = 1289–1310|volume = 103|issue = 8|doi = 10.1109/JPROC.2015.2431914|first1 = You|last1 = Zhou|first2 = S.|last2 = Ramanathan|s2cid = 11347598|url = https://zenodo.org/record/895565}}</ref><ref name=":2">{{Cite document|title=Neuromorphic Circuits With Neural Modulation Enhancing the Information Content of Neural Signaling {{!}} International Conference on Neuromorphic Systems 2020|language=EN|doi=10.1145/3407197.3407204|s2cid=220794387}}</ref> Training software-based neuromorphic systems of spiking neural networks can be achieved using error backpropagation, e.g., using [[Python (programming language)|Python]] based frameworks such as snnTorch,<ref>{{cite journal |last1=Eshraghian|first1=Jason K.|last2=Ward|first2=Max|last3=Neftci |first3=Emre|last4=Wang|first4=Xinxin|last5=Lenz|first5=Gregor|last6=Dwivedi|first6=Girish|last7=Bennamoun|first7=Mohammed|last8=Jeong|first8=Doo Seok|last9=Lu|first9=Wei D.|title=Training Spiking Neural Networks Using Lessons from Deep Learning |date=1 October 2021 |arxiv=2109.12894 }}</ref> or using canonical learning rules from the biological learning literature, e.g., using BindsNet.<ref>{{Cite web | url=https://github.com/Hananel-Hazan/bindsnet | title=Hananel-Hazan/bindsnet: Simulation of spiking neural networks (SNNs) using PyTorch.| date=31 March 2020}}</ref><br />
<br />
Neuromorphic engineering, also known as neuromorphic computing, is the use of very-large-scale integration (VLSI) systems containing electronic analog circuits to mimic neuro-biological architectures present in the nervous system. A neuromorphic computer/chip is any device that uses physical artificial neurons (made from silicon) to do computations. In recent times, the term neuromorphic has been used to describe analog, digital, mixed-mode analog/digital VLSI, and software systems that implement models of neural systems (for perception, motor control, or multisensory integration). The implementation of neuromorphic computing on the hardware level can be realized by oxide-based memristors, spintronic memories, threshold switches, and transistors. Training software-based neuromorphic systems of spiking neural networks can be achieved using error backpropagation, e.g., using Python based frameworks such as snnTorch, or using canonical learning rules from the biological learning literature, e.g., using BindsNet.<br />
<br />
'''<font color="#ff8000">神经形态工程Neuromorphic engineering</font>'''(也称为'''<font color="#ff8000">神经形态计算Neuromorphic computing</font>''')是指使用包含电子'''<font color="#ff8000">模拟电路Analog circuit</font>'''的'''<font color="#ff8000">超大规模集成电路Very-large-scale integration</font>'''系统来模拟神经系统中生理结构的研究方法。神经形态计算机或神经形态芯片包括任何使用由硅制成的人造神经元进行计算的设备。近年来，神经形态学（neuromorphic）这个术语被用来描述能够实现'''<font color="#ff8000">神经系统Neural system</font>'''模型功能(如'''<font color="#ff8000">感知Perception</font>'''、'''<font color="#ff8000">运动控制Motor control</font>'''，'''<font color="#ff8000">多感官整合Multisensory integration</font>'''等)的模拟、数字、模拟/数字混合模式超大规模集成电路和软件系统。神经形态计算的硬件实现可以通过基于氧化物的'''<font color="#ff8000">记忆电阻器Memristor</font>'''、自旋电子存储器、阈值开关和'''<font color="#ff8000">晶体管Transistor</font>'''来实现。对基于软件的脉冲神经网络系统的训练可以通过误差反向传播机制来实现，例如，使用snnTorch等基于Python的框架，或使用BindsNet等典型的受生物启发的学习模式。<br />
<br />
A key aspect of neuromorphic engineering is understanding how the morphology of individual neurons, circuits, applications, and overall architectures creates desirable computations, affects how information is represented, influences robustness to damage, incorporates learning and development, adapts to local change (plasticity), and facilitates evolutionary change.<br />
<br />
A key aspect of neuromorphic engineering is understanding how the morphology of individual neurons, circuits, applications, and overall architectures creates desirable computations, affects how information is represented, influences robustness to damage, incorporates learning and development, adapts to local change (plasticity), and facilitates evolutionary change.<br />
<br />
神经形态工程领域的一个关键问题，就是理解'''<font color="#32CD32">单个神经元形态、神经回路、应用和整体结构</font>'''如何产生理想的计算，如何影响信息的表示和对破坏的鲁棒性，如何整合学习和发展，如何适应局部变化(可塑性) 并促进逐渐发展的变化。<br />
<br />
Neuromorphic engineering is an interdisciplinary subject that takes inspiration from [[biology]], [[physics]], [[mathematics]], [[computer science]], and [[electronic engineering]]<ref name=":2" /> to design artificial neural systems, such as [[Machine vision|vision systems]], head-eye systems, auditory processors, and autonomous robots, whose physical architecture and design principles are based on those of biological nervous systems.<ref>{{Cite journal | doi = 10.1155/2012/705483| title = Qualitative Functional Decomposition Analysis of Evolved Neuromorphic Flight Controllers| journal = Applied Computational Intelligence and Soft Computing| volume = 2012| pages = 1–21| year = 2012| last1 = Boddhu | first1 = S. K. | last2 = Gallagher | first2 = J. C. | doi-access = free}}</ref> It was developed by [[Carver Mead]]<ref>{{cite web|title=carver mead website|last1=Mead|first1=Carver|url=http://www.carvermead.caltech.edu/index.html|website=carvermead}}</ref> in the late 1980s.<br />
<br />
Neuromorphic engineering is an interdisciplinary subject that takes inspiration from biology, physics, mathematics, computer science, and electronic engineering to design artificial neural systems, such as vision systems, head-eye systems, auditory processors, and autonomous robots, whose physical architecture and design principles are based on those of biological nervous systems. It was developed by Carver Mead in the late 1980s.<br />
<br />
神经形态工程是以'''<font color="#ff8000">生物学Biology</font>'''、'''<font color="#ff8000">物理学Physics</font>'''、'''<font color="#ff8000">数学Mathematics</font>'''、'''<font color="#ff8000">计算机科学Computer science</font>'''和'''<font color="#ff8000">电子工程Electronic engineering</font>'''等学科为基础，设计人工神经系统（如'''<font color="#ff8000">视觉系统Vision systems</font>'''、头眼系统、听觉处理器以及物理结构和设计原则都受启发于生物神经系统的自主机器人）的一门交叉学科。20世纪80年代后期，卡弗·米德极大地推动了神经形态工程领域的发展。<br />
<br />
==Neurological inspiration==<br />
Neuromorphic engineering is set apart by the inspiration it takes from what we know about the structure and operations of the [[brain]]. Neuromorphic engineering translates what we know about the brain's function into computer systems. Work has mostly focused on replicating the analog nature of [[biological computation]] and the role of [[neuron]]s in [[cognition]].<br />
<br />
Neuromorphic engineering is set apart by the inspiration it takes from what we know about the structure and operations of the brain. Neuromorphic engineering translates what we know about the brain's function into computer systems. Work has mostly focused on replicating the analog nature of biological computation and the role of neurons in cognition.<br />
<br />
神经学灵感工程学是从我们对大脑结构和运作的了解中获得灵感而形成的。神经形态工程学将我们对大脑功能的了解转化为计算机系统。工作主要集中在复制生物计算的模拟特性和神经元在认知中的作用。<br />
<br />
The biological processes of neurons and their [[synapse]]s are dauntingly complex, and thus very difficult to artificially simulate. A key feature of biological brains is that all of the processing in neurons use analog chemical signals. This makes it hard to replicate brains in computers because the current generation of computers is completely digital. However, the characteristics of these parts can be abstracted into mathematical functions that closely capture the essence of the neuron's operations.<br />
<br />
The biological processes of neurons and their synapses are dauntingly complex, and thus very difficult to artificially simulate. A key feature of biological brains is that all of the processing in neurons use analog chemical signals. This makes it hard to replicate brains in computers because the current generation of computers is completely digital. However, the characteristics of these parts can be abstracted into mathematical functions that closely capture the essence of the neuron's operations.<br />
<br />
神经元及其突触的生物学过程极其复杂，难以人工模拟。生物大脑的一个关键特征是，神经元的所有处理过程都使用模拟化学信号。这使得很难在计算机中复制大脑，因为目前的计算机是完全数字化的。然而，这些部分的特征可以抽象为数学函数，这些函数紧密地抓住了神经元运行的本质。<br />
<br />
The goal of neuromorphic computing is not to perfectly mimic the brain and all of its functions, but instead to extract what is known of its structure and operations to be used in a practical computing system. No neuromorphic system will claim nor attempt to reproduce every element of neurons and synapses, but all adhere to the idea that computation is highly [[distributed processing|distributed]] throughout a series of small computing elements analogous to a neuron. While this sentiment is standard, researchers chase this goal with different methods.<ref>{{Cite journal | doi = 10.1088/1741-2560/13/5/051001| title = Large-scale neuromorphic computing systems| journal = Journal of Neural Engineering| volume = 13| pages = 1–15| year = 2016| last1 = Furber | first1 = Steve| issue = 5| pmid = 27529195| bibcode = 2016JNEng..13e1001F| doi-access = free}}</ref><br />
<br />
The goal of neuromorphic computing is not to perfectly mimic the brain and all of its functions, but instead to extract what is known of its structure and operations to be used in a practical computing system. No neuromorphic system will claim nor attempt to reproduce every element of neurons and synapses, but all adhere to the idea that computation is highly distributed throughout a series of small computing elements analogous to a neuron. While this sentiment is standard, researchers chase this goal with different methods.<br />
<br />
神经形态计算的目标不是完美地模拟大脑及其所有功能，而是提取已知的大脑结构和操作，用于实际的计算系统。没有哪个神经形态学系统会声称或试图复制神经元和突触的每一个元素，但所有人都坚持这样的观点，即计算是高度分布在一系列类似于神经元的小型计算元素中的。虽然这种情绪是标准的，但研究人员用不同的方法追求这一目标。<br />
<br />
==Examples==<br />
As early as 2006, researchers at [[Georgia Tech]] published a field programmable neural array.<ref>{{Cite book|title = A field programmable neural array|last1 = Farquhar|first1 = Ethan|date = May 2006|journal = IEEE International Symposium on Circuits and Systems|pages = 4114–4117|last2 = Hasler|first2 = Paul.|doi = 10.1109/ISCAS.2006.1693534|isbn = 978-0-7803-9389-9|s2cid = 206966013}}</ref> This chip was the first in a line of increasingly complex arrays of floating gate transistors that allowed programmability of charge on the gates of [[MOSFET]]s to model the channel-ion characteristics of neurons in the brain and was one of the first cases of a silicon programmable array of neurons.<br />
<br />
As early as 2006, researchers at Georgia Tech published a field programmable neural array. This chip was the first in a line of increasingly complex arrays of floating gate transistors that allowed programmability of charge on the gates of MOSFETs to model the channel-ion characteristics of neurons in the brain and was one of the first cases of a silicon programmable array of neurons.<br />
<br />
早在2006年，佐治亚理工学院的研究人员就发表了现场可编程神经阵列。这种芯片是一系列越来越复杂的浮栅晶体管阵列中的第一个，这些晶体管可以在 mosfet 的栅极上编程来模拟大脑中神经元的通道离子特性，这也是硅可编程神经元阵列的首批案例之一。<br />
<br />
In November 2011, a group of [[MIT]] researchers created a computer chip that mimics the analog, ion-based communication in a synapse between two neurons using 400 transistors and standard [[CMOS]] manufacturing techniques.<ref>{{cite web|title=MIT creates "brain chip"|url=http://www.extremetech.com/extreme/105067-mit-creates-brain-chip|access-date=4 December 2012}}</ref><ref name="Neuromorphic silicon paper">{{cite journal|title=Neuromorphic silicon neurons and large-scale neural networks: challenges and opportunities|doi=10.3389/fnins.2011.00108|pmid=21991244|pmc=3181466|volume=5|pages=108|journal=Frontiers in Neuroscience|year=2011|last1=Poon|first1=Chi-Sang|last2=Zhou|first2=Kuan|doi-access=free}}</ref><br />
<br />
In November 2011, a group of MIT researchers created a computer chip that mimics the analog, ion-based communication in a synapse between two neurons using 400 transistors and standard CMOS manufacturing techniques.<br />
<br />
2011年11月，麻省理工学院的一组研究人员发明了一种计算机芯片，该芯片使用400个晶体管和标准的 CMOS 制造技术，在两个神经元之间的突触中模拟模拟基于离子的通讯。<br />
<br />
In June 2012, [[spintronic]] researchers at [[Purdue University]] presented a paper on the design of a neuromorphic chip using [[Spin valve|lateral spin valve]]s and [[memristor]]s. They argue that the architecture works similarly to neurons and can therefore be used to test methods of reproducing the brain's processing. In addition, these chips are significantly more energy-efficient than conventional ones.<ref name="Spin Devices Prop">{{Cite arXiv|title=Proposal For Neuromorphic Hardware Using Spin Devices|eprint=1206.3227|last1=Sharad|first1=Mrigank|last2=Augustine|first2=Charles|last3=Panagopoulos|first3=Georgios|last4=Roy|first4=Kaushik|class=cond-mat.dis-nn|year=2012}}</ref><br />
<br />
In June 2012, spintronic researchers at Purdue University presented a paper on the design of a neuromorphic chip using lateral spin valves and memristors. They argue that the architecture works similarly to neurons and can therefore be used to test methods of reproducing the brain's processing. In addition, these chips are significantly more energy-efficient than conventional ones.<br />
<br />
2012年6月，普渡大学的自旋电子学研究人员发表了一篇关于利用侧向自旋阀和记忆电阻器设计神经形态芯片的论文。他们认为，这种结构的工作原理与神经元相似，因此可以用来测试复制大脑处理过程的方法。此外，这些芯片明显比传统芯片更节能。<br />
<br />
Research at [[HP Labs]] on Mott memristors has shown that while they can be non-[[Volatile memory|volatile]], the volatile behavior exhibited at temperatures significantly below the [[phase transition]] temperature can be exploited to fabricate a [[neuristor]],<ref name=":0" /> a biologically-inspired device that mimics behavior found in neurons.<ref name=":0">{{Cite journal | doi = 10.1038/nmat3510| pmid = 23241533| title = A scalable neuristor built with Mott memristors| journal = Nature Materials| volume = 12| issue = 2| pages = 114–7| year = 2012| last1 = Pickett | first1 = M. D. | last2 = Medeiros-Ribeiro | first2 = G. | last3 = Williams | first3 = R. S. | bibcode = 2013NatMa..12..114P| s2cid = 16271627| url = https://semanticscholar.org/paper/b6ba6f496ace2b947f111059663e76bb60e9efeb}}</ref> In September 2013, they presented models and simulations that show how the spiking behavior of these neuristors can be used to form the components required for a [[Turing machine]].<ref>{{cite journal|doi=10.1088/0957-4484/24/38/384002|title=Phase transitions enable computational universality in neuristor-based cellular automata|author1=Matthew D Pickett|author2=R Stanley Williams|name-list-style=amp|date=September 2013|publisher=IOP Publishing Ltd|journal=Nanotechnology|volume=24|issue=38|pmid=23999059|bibcode=2013Nanot..24L4002P|at=384002}}</ref><br />
<br />
Research at HP Labs on Mott memristors has shown that while they can be non-volatile, the volatile behavior exhibited at temperatures significantly below the phase transition temperature can be exploited to fabricate a neuristor, a biologically-inspired device that mimics behavior found in neurons. In September 2013, they presented models and simulations that show how the spiking behavior of these neuristors can be used to form the components required for a Turing machine.<br />
<br />
惠普实验室在莫特记忆电阻器上的研究表明，尽管它们可以是非挥发性的，但是在相变温度以下显示的挥发性行为可以被用来制造神经元电阻器，这是一种模仿神经元行为的生物设备。2013年9月，他们展示了模型和仿真，展示了这些神经元的尖峰行为是如何被用来形成图灵机所需的元件的。<br />
<br />
[[Neurogrid]], built by ''Brains in Silicon'' at [[Stanford University]],<ref>{{cite journal|last1=Boahen|first1=Kwabena|title=Neurogrid: A Mixed-Analog-Digital Multichip System for Large-Scale Neural Simulations|journal=Proceedings of the IEEE|date=24 April 2014|volume=102|issue=5|pages=699–716|doi=10.1109/JPROC.2014.2313565|s2cid=17176371}}</ref> is an example of hardware designed using neuromorphic engineering principles. The circuit board is composed of 16 custom-designed chips, referred to as NeuroCores. Each NeuroCore's analog circuitry is designed to emulate neural elements for 65536 neurons, maximizing energy efficiency. The emulated neurons are connected using digital circuitry designed to maximize spiking throughput.<ref>{{cite journal|doi=10.1038/503022a|pmid = 24201264|title = Neuroelectronics: Smart connections|journal = Nature|volume = 503|issue = 7474|pages = 22–4|year = 2013|last1 = Waldrop|first1 = M. Mitchell|bibcode = 2013Natur.503...22W|doi-access = free}}</ref><ref>{{cite journal|doi=10.1109/JPROC.2014.2313565|title = Neurogrid: A Mixed-Analog-Digital Multichip System for Large-Scale Neural Simulations|journal = Proceedings of the IEEE|volume = 102|issue = 5|pages = 699–716|year = 2014|last1 = Benjamin|first1 = Ben Varkey|last2 = Peiran Gao|last3 = McQuinn|first3 = Emmett|last4 = Choudhary|first4 = Swadesh|last5 = Chandrasekaran|first5 = Anand R.|last6 = Bussat|first6 = Jean-Marie|last7 = Alvarez-Icaza|first7 = Rodrigo|last8 = Arthur|first8 = John V.|last9 = Merolla|first9 = Paul A.|last10 = Boahen|first10 = Kwabena|s2cid = 17176371}}</ref><br />
<br />
Neurogrid, built by Brains in Silicon at Stanford University, is an example of hardware designed using neuromorphic engineering principles. The circuit board is composed of 16 custom-designed chips, referred to as NeuroCores. Each NeuroCore's analog circuitry is designed to emulate neural elements for 65536 neurons, maximizing energy efficiency. The emulated neurons are connected using digital circuitry designed to maximize spiking throughput.<br />
<br />
神经网格是由斯坦福大学硅谷大脑公司建立的，是一个使用神经形态工程原理设计硬件的例子。该电路板由16个定制设计的芯片组成，称为 NeuroCores。每个 NeuroCore 的模拟电路被设计为模拟65536个神经元的神经元元件，最大限度地提高能量效率。模拟的神经元通过设计的数字电路连接，以最大化脉冲吞吐量。<br />
<br />
A research project with implications for neuromorphic engineering is the [[Human Brain Project]] that is attempting to simulate a complete human brain in a supercomputer using biological data. It is made up of a group of researchers in neuroscience, medicine, and computing.<ref>{{cite web|title=Involved Organizations|url=http://www.humanbrainproject.eu/partners.html|access-date=22 February 2013|url-status=dead|archive-url=https://web.archive.org/web/20130302142627/http://www.humanbrainproject.eu/partners.html|archive-date=2 March 2013}}</ref> [[Henry Markram]], the project's co-director, has stated that the project proposes to establish a foundation to explore and understand the brain and its diseases, and to use that knowledge to build new computing technologies. The three primary goals of the project are to better understand how the pieces of the brain fit and work together, to understand how to objectively diagnose and treat brain diseases, and to use the understanding of the human brain to develop neuromorphic computers. That the simulation of a complete human brain will require a supercomputer a thousand times more powerful than today's encourages the current focus on neuromorphic computers.<ref>{{cite web|title=Human Brain Project|url=http://www.humanbrainproject.eu|access-date=22 February 2013}}</ref> $1.3 billion has been allocated to the project by The [[European Commission]].<ref>{{cite web|title=The Human Brain Project and Recruiting More Cyberwarriors|url=http://www.marketplace.org/topics/tech/human-brain-project-and-recruiting-more-cyberwarriors|access-date=22 February 2013|date=January 29, 2013}}</ref><br />
<br />
A research project with implications for neuromorphic engineering is the Human Brain Project that is attempting to simulate a complete human brain in a supercomputer using biological data. It is made up of a group of researchers in neuroscience, medicine, and computing. Henry Markram, the project's co-director, has stated that the project proposes to establish a foundation to explore and understand the brain and its diseases, and to use that knowledge to build new computing technologies. The three primary goals of the project are to better understand how the pieces of the brain fit and work together, to understand how to objectively diagnose and treat brain diseases, and to use the understanding of the human brain to develop neuromorphic computers. That the simulation of a complete human brain will require a supercomputer a thousand times more powerful than today's encourages the current focus on neuromorphic computers. $1.3 billion has been allocated to the project by The European Commission.<br />
<br />
一个对神经形态工程有影响的研究项目是人脑项目，它试图用生物数据在超级计算机中模拟完整的人脑。它由一群神经科学、医学和计算机科学的研究人员组成。该项目的联合主管亨利•马克拉姆(Henry Markram)表示，该项目提议建立一个基础，以探索和了解大脑及其疾病，并利用这些知识构建新的计算机技术。这个项目的三个主要目标是: 更好地理解大脑的各个部分是如何相互匹配和协同工作的; 理解如何客观地诊断和治疗脑部疾病; 以及利用对人类大脑的理解来开发神经形态计算机。模拟一个完整的人类大脑需要一台比现在强大一千倍的超级计算机，这鼓励了当前对神经形态计算机的关注。欧洲委员会已经拨款13亿美元用于这个项目。<br />
<br />
Other research with implications for neuromorphic engineering involves the [[BRAIN Initiative]]<ref name="economist">[https://www.economist.com/news/science-and-technology/21582495-computers-will-help-people-understand-brains-better-and-understanding-brains Neuromorphic computing: The machine of a new soul], The Economist, 2013-08-03</ref> and the [[TrueNorth]] chip from [[IBM]].<ref>{{cite journal|last1=Modha|first1=Dharmendra|title=A million spiking-neuron integrated circuit with a scalable communication network and interface|journal=Science|date=Aug 2014|volume=345|issue=6197|pages=668–673|doi=10.1126/science.1254642|pmid=25104385|bibcode=2014Sci...345..668M|s2cid=12706847}}</ref> Neuromorphic devices have also been demonstrated using nanocrystals, nanowires, and conducting polymers.<ref>{{Cite web|url=http://jessamynfairfield.com/wp-content/uploads/2017/03/PWMar17Fairfield.pdf|title=Smarter Machines|last=Fairfield|first=Jessamyn|date=March 1, 2017}}</ref><br />
<br />
Other research with implications for neuromorphic engineering involves the BRAIN InitiativeNeuromorphic computing: The machine of a new soul, The Economist, 2013-08-03 and the TrueNorth chip from IBM. Neuromorphic devices have also been demonstrated using nanocrystals, nanowires, and conducting polymers.<br />
<br />
其他与神经形态工程有关的研究还包括 BRAIN initiativity/euromorphic computing: The machine of a new soul，The Economist，2013-08-03 and The TrueNorth chip from IBM。神经形态学设备也已经被证明使用纳米晶体、纳米线和导电聚合物。<br />
<br />
[[Intel]] unveiled its neuromorphic research chip, called “[[Intel Loihi|Loihi]]”, in October 2017. The chip uses an asynchronous [[spiking neural network]] (SNN) to implement adaptive self-modifying event-driven fine-grained parallel computations used to implement learning and inference with high efficiency.<ref>{{cite journal |last1=Davies |first1=Mike |title=Loihi: A Neuromorphic Manycore Processor with On-Chip Learning |journal=IEEE Micro |date=January 16, 2018 |volume=38 |issue=1 |pages=82–99 |display-authors=etal|doi=10.1109/MM.2018.112130359 |s2cid=3608458 }}</ref><ref name="Morris2017">{{cite web |last1=Morris |first1=John |title=Why Intel built a neuromorphic chip |url=https://www.zdnet.com/article/why-intel-built-a-neuromorphic-chip/ |website=ZDNet |access-date=17 August 2018 |language=en}}</ref><br />
<br />
Intel unveiled its neuromorphic research chip, called “Loihi”, in October 2017. The chip uses an asynchronous spiking neural network (SNN) to implement adaptive self-modifying event-driven fine-grained parallel computations used to implement learning and inference with high efficiency.<br />
<br />
2017年10月，英特尔公布了其神经形态研究芯片“ Loihi”。该芯片采用异步脉冲神经网络(SNN)实现自适应自修改事件驱动的细粒度并行计算，实现了高效的学习和推理。<br />
<br />
[[IMEC]], a Belgium-based nanoelectronics research center, demonstrated the world's first self-learning neuromorphic chip. The brain-inspired chip, based on OxRAM technology, has the capability of self-learning and has been demonstrated to have the ability to compose music.<ref>{{cite web |title=Imec demonstrates self-learning neuromorphic chip that composes music |url=https://www.imec-int.com/en/articles/imec-demonstrates-self-learning-neuromorphic-chip-that-composes-music |website=IMEC International |access-date=1 October 2019}}</ref> IMEC released the 30-second tune composed by the prototype. The chip was sequentially loaded with songs in the same time signature and style. The songs were old Belgian and French flute minuets, from which the chip learned the rules at play and then applied them.<ref>{{cite web|last1=Bourzac|first1=Katherine|title=A Neuromorphic Chip That Makes Music|url=https://spectrum.ieee.org/a-neuromorphic-chip-that-makes-music|url-status=live|access-date=1 October 2019|website=IEEE Spectrum|date=May 23, 2017}}</ref><br />
<br />
IMEC, a Belgium-based nanoelectronics research center, demonstrated the world's first self-learning neuromorphic chip. The brain-inspired chip, based on OxRAM technology, has the capability of self-learning and has been demonstrated to have the ability to compose music. IMEC released the 30-second tune composed by the prototype. The chip was sequentially loaded with songs in the same time signature and style. The songs were old Belgian and French flute minuets, from which the chip learned the rules at play and then applied them.<br />
<br />
比利时的一个纳米电子学研究中心 IMEC 展示了世界上第一个自学习神经形态芯片。这种基于 OxRAM 技术的大脑启发芯片具有自学习能力，并已被证明具有创作音乐的能力。IMEC 发布了由原型机谱写的30秒曲调。芯片按顺序加载同时签名和风格的歌曲。歌曲是古老的比利时和法国长笛小步舞曲，筹码从中学习游戏规则，然后应用它们。<br />
<br />
[[Blue Brain Project|The Blue Brain Project]], led by Henry Markram, aims to build biologically detailed digital reconstructions and simulations of the mouse brain. The Blue Brain Project has created in silico models of rodent brains, while attempting to replicate as many details about its biology as possible. The supercomputer-based simulations offer new perspectives on understanding the structure and functions of the brain.<br />
<br />
The Blue Brain Project, led by Henry Markram, aims to build biologically detailed digital reconstructions and simulations of the mouse brain. The Blue Brain Project has created in silico models of rodent brains, while attempting to replicate as many details about its biology as possible. The supercomputer-based simulations offer new perspectives on understanding the structure and functions of the brain.<br />
<br />
由亨利 · 马克拉姆领导的蓝脑计划旨在建立小鼠大脑的生物细节数字重建和模拟。蓝色大脑项目已经在啮齿动物大脑的电子模型中创建，同时试图尽可能多地复制其生物学细节。基于超级计算机的模拟为理解大脑的结构和功能提供了新的视角。<br />
<br />
The European Union funded a series of projects at the University of Heidelberg, which led to the development of <br />
[[BrainScaleS]] (brain-inspired multiscale computation in neuromorphic hybrid systems), a hybrid analog [[neuromorphic]] supercomputer located at Heidelberg University, Germany. It was developed as part of the [[Human Brain Project]] neuromorphic computing platform and is the complement to the [[SpiNNaker]] supercomputer (which is based on digital technology). The architecture used in BrainScaleS mimics biological neurons and their connections on a physical level; additionally, since the components are made of silicon, these model neurons operate on average 864 times (24 hours of real time is 100 seconds in the machine simulation) that of their biological counterparts.<ref>{{Cite web|date=2016-03-21|title=Beyond von Neumann, Neuromorphic Computing Steadily Advances|url=https://www.hpcwire.com/2016/03/21/lacking-breakthrough-neuromorphic-computing-steadily-advance/|access-date=2021-10-08|website=HPCwire|language=en-US}}</ref><br />
<br />
The European Union funded a series of projects at the University of Heidelberg, which led to the development of <br />
BrainScaleS (brain-inspired multiscale computation in neuromorphic hybrid systems), a hybrid analog neuromorphic supercomputer located at Heidelberg University, Germany. It was developed as part of the Human Brain Project neuromorphic computing platform and is the complement to the SpiNNaker supercomputer (which is based on digital technology). The architecture used in BrainScaleS mimics biological neurons and their connections on a physical level; additionally, since the components are made of silicon, these model neurons operate on average 864 times (24 hours of real time is 100 seconds in the machine simulation) that of their biological counterparts.<br />
<br />
欧盟资助了海德堡大学的一系列项目，这些项目导致了 BrainScaleS (神经形态混合系统中受大脑启发的多尺度计算)的发展，这是一台位于德国海德堡大学的混合模拟神经形态超级计算机。它是作为人脑计划神经形态计算平台的一部分而开发的，是 SpiNNaker 超级计算机(基于数字技术)的补充。大脑尺度中使用的体系结构模拟了生物神经元及其在物理层面上的连接; 此外，由于这些组件是由硅制成的，这些模型神经元平均运行864倍(在机器模拟中，24小时的实时时间是100秒)。<br />
<br />
===Neuromorphic sensors===<br />
The concept of neuromorphic systems can be extended to sensors (not just to computation). An example of this applied to detecting [[light]] is the [[retinomorphic sensor]] or, when employed in an array, the [[event camera]].<br />
<br />
The concept of neuromorphic systems can be extended to sensors (not just to computation). An example of this applied to detecting light is the retinomorphic sensor or, when employed in an array, the event camera.<br />
<br />
= = = 神经形态传感器 = = = 神经形态系统的概念可以扩展到传感器(而不仅仅是计算)。用于检测光线的一个例子是视网膜变形传感器，或者在阵列中使用的事件摄像机。<br />
<br />
==Ethical considerations==<br />
While the interdisciplinary concept of neuromorphic engineering is relatively new, many of the same ethical considerations apply to neuromorphic systems as apply to [[human-like machines]] and [[artificial intelligence]] in general. However, the fact that neuromorphic systems are designed to mimic a [[human brain]] gives rise to unique ethical questions surrounding their usage.<br />
<br />
While the interdisciplinary concept of neuromorphic engineering is relatively new, many of the same ethical considerations apply to neuromorphic systems as apply to human-like machines and artificial intelligence in general. However, the fact that neuromorphic systems are designed to mimic a human brain gives rise to unique ethical questions surrounding their usage.<br />
<br />
虽然神经形态工程的跨学科概念相对较新，但许多同样的伦理考虑适用于神经形态系统，就像适用于类人机器和一般人工智能一样。然而，神经形态系统是为了模仿人类大脑而设计的这一事实引起了围绕其使用的独特的伦理问题。<br />
<br />
However, the practical debate is that neuromorphic hardware as well as artificial "neural networks" are immensely simplified models of how the brain operates or processes information at a much lower [[complex system|complexity in terms of size and functional technology]] and a much more regular structure in terms of [[brain connectivity|connectivity]]. Comparing [[neuromorphic chip]]s to the brain is a very crude comparison similar to comparing a plane to a bird just because they both have wings and a tail. The fact is that neural cognitive systems are many orders of magnitude more [[Energy efficiency (physics)|energy-]] and compute-efficient than current state-of-the-art AI and neuromorphic engineering is an attempt to narrow this gap by inspiring from the brain's mechanism just like many engineering designs have [[bioengineering|bio-inspired features]].<br />
<br />
However, the practical debate is that neuromorphic hardware as well as artificial "neural networks" are immensely simplified models of how the brain operates or processes information at a much lower complexity in terms of size and functional technology and a much more regular structure in terms of connectivity. Comparing neuromorphic chips to the brain is a very crude comparison similar to comparing a plane to a bird just because they both have wings and a tail. The fact is that neural cognitive systems are many orders of magnitude more energy- and compute-efficient than current state-of-the-art AI and neuromorphic engineering is an attempt to narrow this gap by inspiring from the brain's mechanism just like many engineering designs have bio-inspired features.<br />
<br />
然而，实际上的争论是，神经形态硬件和人工“神经网络”是大脑如何运作或处理信息的极其简化的模型，在大小和功能技术方面的复杂性要低得多，在连接方面的结构则更加规则。将神经形态芯片与大脑进行比较是一种非常粗糙的比较，类似于仅仅因为一架飞机有翅膀和一条尾巴就将它与一只鸟进行比较。事实上，神经认知系统比当前最先进的人工智能具有更多的能量和计算效率，而神经形态工程是一种通过从大脑机制中激发灵感来缩小这种差距的尝试，就像许多工程设计具有生物启发的特征一样。数量级。<br />
<br />
===Democratic concerns===<br />
Significant ethical limitations may be placed on neuromorphic engineering due to public perception.<ref>{{Cite report|url=https://ai100.stanford.edu/sites/g/files/sbiybj9861/f/ai_100_report_0831fnl.pdf|title=Artificial Intelligence and Life in 2030|author=2015 Study Panel|date=September 2016|work=One Hundred Year Study on Artificial Intelligence (AI100)|publisher=Stanford University}}</ref> Special [[Eurobarometer]] 382: Public Attitudes Towards Robots, a survey conducted by the European Commission, found that 60% of [[European Union]] citizens wanted a ban of robots in the care of children, the elderly, or the disabled. Furthermore, 34% were in favor of a ban on robots in education, 27% in healthcare, and 20% in leisure. The European Commission classifies these areas as notably “human.” The report cites increased public concern with robots that are able to mimic or replicate human functions. Neuromorphic engineering, by definition, is designed to replicate the function of the human brain.<ref name=":1">{{Cite web|url=http://ec.europa.eu/commfrontoffice/publicopinion/archives/ebs/ebs_382_en.pdf|title=Special Eurobarometer 382: Public Attitudes Towards Robots|last=European Commission|date=September 2012|website=European Commission}}</ref><br />
<br />
Significant ethical limitations may be placed on neuromorphic engineering due to public perception. Special Eurobarometer 382: Public Attitudes Towards Robots, a survey conducted by the European Commission, found that 60% of European Union citizens wanted a ban of robots in the care of children, the elderly, or the disabled. Furthermore, 34% were in favor of a ban on robots in education, 27% in healthcare, and 20% in leisure. The European Commission classifies these areas as notably “human.” The report cites increased public concern with robots that are able to mimic or replicate human functions. Neuromorphic engineering, by definition, is designed to replicate the function of the human brain.<br />
<br />
由于公众的认知，神经形态工程学可能会受到严重的伦理限制。欧盟委员会进行的一项调查发现，60% 的欧盟公民希望禁止照顾儿童、老人或残疾人的机器人。此外，34% 的人支持禁止机器人用于教育，27% 的人支持医疗保健，20% 的人支持休闲。欧盟委员会将这些地区明显归类为“人类”报告指出，公众越来越关注能够模仿或复制人类功能的机器人。神经形态工程，顾名思义，是为了复制人脑的功能而设计的。<br />
<br />
The democratic concerns surrounding neuromorphic engineering are likely to become even more profound in the future. The European Commission found that EU citizens between the ages of 15 and 24 are more likely to think of robots as human-like (as opposed to instrument-like) than EU citizens over the age of 55. When presented an image of a robot that had been defined as human-like, 75% of EU citizens aged 15–24 said it corresponded with the idea they had of robots while only 57% of EU citizens over the age of 55 responded the same way. The human-like nature of neuromorphic systems, therefore, could place them in the categories of robots many EU citizens would like to see banned in the future.<ref name=":1" /><br />
<br />
The democratic concerns surrounding neuromorphic engineering are likely to become even more profound in the future. The European Commission found that EU citizens between the ages of 15 and 24 are more likely to think of robots as human-like (as opposed to instrument-like) than EU citizens over the age of 55. When presented an image of a robot that had been defined as human-like, 75% of EU citizens aged 15–24 said it corresponded with the idea they had of robots while only 57% of EU citizens over the age of 55 responded the same way. The human-like nature of neuromorphic systems, therefore, could place them in the categories of robots many EU citizens would like to see banned in the future.<br />
<br />
围绕神经形态工程的民主关注可能在未来变得更加深刻。欧盟委员会(European Commission)发现，15至24岁的欧盟公民比55岁以上的欧盟公民更有可能认为机器人像人(而不是像仪器)。当看到一张被定义为“类人”的机器人图片时，年龄在15岁至24岁之间的欧盟公民中有75% 的人表示，这与他们对机器人的想法相符，而55岁以上的欧盟公民中只有57% 的人有同样的反应。因此，类似人类的神经形态系统，可以把它们归入许多欧盟公民希望在未来禁止使用的机器人类别。<br />
<br />
===Personhood===<br />
As neuromorphic systems have become increasingly advanced, some scholars{{who|date=August 2021}} have advocated for granting [[personhood]] rights to these systems. If the brain is what grants humans their personhood, to what extent does a neuromorphic system have to mimic the human brain to be granted personhood rights? Critics of technology development in the [[Human Brain Project]], which aims to advance brain-inspired computing, have argued that advancement in neuromorphic computing could lead to machine consciousness or personhood.<ref>{{Cite journal|last=Aicardi|first=Christine|date=September 2018|title=Accompanying technology development in the Human Brain Project: From foresight to ethics management|journal=Futures|volume=102|pages=114–124|doi=10.1016/j.futures.2018.01.005|doi-access=free}}</ref> If these systems are to be treated as people, critics argue, then many tasks humans perform using neuromorphic systems, including the act of termination of neuromorphic systems, may be morally impermissible as these acts would violate the autonomy of the neuromorphic systems.<ref>{{Cite journal|last=Lim|first=Daniel|date=2014-06-01|title=Brain simulation and personhood: a concern with the Human Brain Project|journal=Ethics and Information Technology|language=en|volume=16|issue=2|pages=77–89|doi=10.1007/s10676-013-9330-5|s2cid=17415814|issn=1572-8439}}</ref><br />
<br />
As neuromorphic systems have become increasingly advanced, some scholars have advocated for granting personhood rights to these systems. If the brain is what grants humans their personhood, to what extent does a neuromorphic system have to mimic the human brain to be granted personhood rights? Critics of technology development in the Human Brain Project, which aims to advance brain-inspired computing, have argued that advancement in neuromorphic computing could lead to machine consciousness or personhood. If these systems are to be treated as people, critics argue, then many tasks humans perform using neuromorphic systems, including the act of termination of neuromorphic systems, may be morally impermissible as these acts would violate the autonomy of the neuromorphic systems.<br />
<br />
= = = 人格 = = 随着神经形态系统的日益发展，一些学者主张赋予这些系统人格权。如果是大脑赋予了人类人格，那么神经形态系统在多大程度上必须模仿人类大脑才能被赋予人格权利？“人脑计划”旨在推进以大脑为灵感的计算机技术发展，该计划的批评者认为，神经形态计算机技术的进步可能导致机器意识或人格的形成。批评家认为，如果这些系统被当作人来对待，那么人类使用神经形态系统执行的许多任务，包括终止神经形态系统的行为，可能在道德上是不允许的，因为这些行为将违反神经形态系统的自主性。<br />
<br />
==Dual use (military applications)==<br />
The [[Joint Artificial Intelligence Center]], a branch of the U.S. military, is a center dedicated to the procurement and implementation of AI software and neuromorphic hardware for combat use. Specific applications include smart headsets/goggles and robots. JAIC intends to rely heavily on neuromorphic technology to connect "every fighter every shooter" within a network of neuromorphic-enabled units.<br />
<br />
The Joint Artificial Intelligence Center, a branch of the U.S. military, is a center dedicated to the procurement and implementation of AI software and neuromorphic hardware for combat use. Specific applications include smart headsets/goggles and robots. JAIC intends to rely heavily on neuromorphic technology to connect "every fighter every shooter" within a network of neuromorphic-enabled units.<br />
<br />
军民两用联合人工智能中心是美国军队的一个分支，专门从事采购和实施用于战斗的人工智能软件和神经形态硬件。具体应用包括智能耳机/护目镜和机器人。JAIC 打算严重依赖神经形态技术来连接神经形态单位网络中的“每个战士每个射手”。<br />
<br />
==Legal considerations==<br />
Skeptics have argued that there is no way to apply the electronic personhood, the concept of personhood that would apply to neuromorphic technology, legally. In a letter signed by 285 experts in law, robotics, medicine, and ethics opposing a European Commission proposal to recognize “smart robots” as legal persons, the authors write, “A legal status for a robot can’t derive from the [[Natural person|Natural Person]] model, since the robot would then hold [[human rights]], such as the right to dignity, the right to its integrity, the right to remuneration or the right to citizenship, thus directly confronting the Human rights. This would be in contradiction with the [[Charter of Fundamental Rights of the European Union]] and the [[Convention for the Protection of Human Rights and Fundamental Freedoms]].”<ref>{{Cite web|url=http://www.robotics-openletter.eu/|title=Robotics Openletter {{!}} Open letter to the European Commission|language=fr-FR|access-date=2019-05-10}}</ref><br />
<br />
Skeptics have argued that there is no way to apply the electronic personhood, the concept of personhood that would apply to neuromorphic technology, legally. In a letter signed by 285 experts in law, robotics, medicine, and ethics opposing a European Commission proposal to recognize “smart robots” as legal persons, the authors write, “A legal status for a robot can’t derive from the Natural Person model, since the robot would then hold human rights, such as the right to dignity, the right to its integrity, the right to remuneration or the right to citizenship, thus directly confronting the Human rights. This would be in contradiction with the Charter of Fundamental Rights of the European Union and the Convention for the Protection of Human Rights and Fundamental Freedoms.”<br />
<br />
= = 法律方面的考虑 = = 怀疑论者认为，在法律上没有办法应用电子人格，这个人格概念将适用于神经形态技术。在一封由285名法律、机器人、医学和伦理学专家签名的信中，作者们反对欧盟委员会承认“智能机器人”为法人的提议，他们写道，“机器人的法律地位不能从自然人模型中推导出来，因为机器人将拥有人权，如尊严权、完整权、报酬权或公民权，从而直接面对人权。这将有悖于《欧洲联盟基本权利宪章和《保护人权和基本自由公约》。”<br />
<br />
=== Ownership and property rights===<br />
There is significant legal debate around property rights and artificial intelligence. In ''Acohs Pty Ltd v. Ucorp Pty Ltd'', Justice Christopher Jessup of the [[Federal Court of Australia]] found that the [[source code]] for [[Material safety data sheets|Material Safety Data Sheets]] could not be [[Copyright law of Australia|copyrighted]] as it was generated by a [[software interface]] rather than a human author.<ref>{{Cite web|url=http://www.lavan.com.au/advice/intellectual_property_technology/copyright_in_source_code_and_digital_products|title=Copyright in source code and digital products|last=Lavan|website=Lavan|language=en|access-date=2019-05-10}}</ref> The same question may apply to neuromorphic systems: if a neuromorphic system successfully mimics a human brain and produces a piece of original work, who, if anyone, should be able to claim ownership of the work?<ref>{{cite journal |last1=Eshraghian|first1=Jason K. |title=Human Ownership of Artificial Creativity |journal=Nature Machine Intelligence |date=9 March 2020 |volume=2 |pages=157–160 |doi=10.1038/s42256-020-0161-x}}</ref><br />
<br />
There is significant legal debate around property rights and artificial intelligence. In Acohs Pty Ltd v. Ucorp Pty Ltd, Justice Christopher Jessup of the Federal Court of Australia found that the source code for Material Safety Data Sheets could not be copyrighted as it was generated by a software interface rather than a human author. The same question may apply to neuromorphic systems: if a neuromorphic system successfully mimics a human brain and produces a piece of original work, who, if anyone, should be able to claim ownership of the work?<br />
<br />
= = = 所有权和财产权 = = = 围绕财产权和人工智能有着重大的法律争论。在 Acohs Pty Ltd 诉 Ucorp Pty Ltd 一案中，澳大利亚联邦法院的克里斯托弗 · 杰瑟普法官发现，材料安全数据表的源代码不能受版权保护，因为它是由软件界面而不是人工作者生成的。同样的问题可能也适用于神经形态系统: 如果一个神经形态系统成功地模仿了人类的大脑并产生了一部原创作品，那么谁，如果有人，应该声称拥有这部作品的所有权？<br />
<br />
==Neuromemristive systems==<br />
Neuromemristive systems are a subclass of neuromorphic computing systems that focus on the use of [[memristors]] to implement [[neuroplasticity]]. While neuromorphic engineering focuses on mimicking biological behavior, neuromemristive systems focus on abstraction.<ref>{{Cite web|url=https://digitalops.sandia.gov/Mediasite/Play/a10cf6ceb55d47608bb8326dd00e46611d|title=002.08 N.I.C.E. Workshop 2014: Towards Intelligent Computing with Neuromemristive Circuits and Systems - Feb. 2014|website=digitalops.sandia.gov|access-date=2019-08-26}}</ref> For example, a neuromemristive system may replace the details of a [[Cerebral cortex|cortical]] microcircuit's behavior with an abstract neural network model.<ref>C. Merkel and D. Kudithipudi, "Neuromemristive extreme learning machines for pattern classification," ISVLSI, 2014.</ref><br />
<br />
Neuromemristive systems are a subclass of neuromorphic computing systems that focus on the use of memristors to implement neuroplasticity. While neuromorphic engineering focuses on mimicking biological behavior, neuromemristive systems focus on abstraction. For example, a neuromemristive system may replace the details of a cortical microcircuit's behavior with an abstract neural network model.C. Merkel and D. Kudithipudi, "Neuromemristive extreme learning machines for pattern classification," ISVLSI, 2014.<br />
<br />
神经记忆电阻系统是神经形态计算系统的一个亚类，主要研究利用记忆电阻器实现神经可塑性。神经形态工程的重点是模拟生物行为，而神经记忆电阻系统的重点是提取。例如，一个神经记忆系统可以用一个抽象的神经网络模型取代皮层微电路的行为细节。默克尔和 d. Kudithipudi，“用于模式分类的神经记忆极端学习机器”，ISVLSI，2014。<br />
<br />
There exist several neuron inspired threshold logic functions<ref name="Maan 1–13" /> implemented with memristors that have applications in high level [[pattern recognition]] applications. Some of the applications reported recently include [[speech recognition]],<ref>{{Cite journal|title = Memristor pattern recogniser: isolated speech word recognition|journal = Electronics Letters|pages = 1370–1372|volume = 51|issue = 17|doi = 10.1049/el.2015.1428|first1 = A.K.|last1 = Maan|first2 = A.P.|last2 = James|first3 = S.|last3 = Dimitrijev|year = 2015|bibcode = 2015ElL....51.1370M|hdl = 10072/140989|s2cid = 61454815|url = https://semanticscholar.org/paper/48d3ab11ec6e213b62f11eedcfb7b7febb058674|hdl-access = free}}</ref> [[face recognition]]<ref>{{Cite journal|title = Memristive Threshold Logic Face Recognition|journal = Procedia Computer Science|date = 2014-01-01|pages = 98–103|volume = 41|series = 5th Annual International Conference on Biologically Inspired Cognitive Architectures, 2014 BICA|doi = 10.1016/j.procs.2014.11.090|first1 = Akshay Kumar|last1 = Maan|first2 = Dinesh S.|last2 = Kumar|first3 = Alex Pappachen|last3 = James|doi-access = free}}</ref> and [[object recognition]].<ref>{{Cite journal|title = Memristive Threshold Logic Circuit Design of Fast Moving Object Detection|journal = IEEE Transactions on Very Large Scale Integration (VLSI) Systems|date = 2015-10-01|issn = 1063-8210|pages = 2337–2341|volume = 23|issue = 10|doi = 10.1109/TVLSI.2014.2359801|first1 = A.K.|last1 = Maan|first2 = D.S.|last2 = Kumar|first3 = S.|last3 = Sugathan|first4 = A.P.|last4 = James|arxiv = 1410.1267|s2cid = 9647290}}</ref> They also find applications in replacing conventional digital logic gates.<ref>{{Cite journal|title = Resistive Threshold Logic|journal = IEEE Transactions on Very Large Scale Integration (VLSI) Systems|date = 2014-01-01|issn = 1063-8210|pages = 190–195|volume = 22|issue = 1|doi = 10.1109/TVLSI.2012.2232946|first1 = A.P.|last1 = James|first2 = L.R.V.J.|last2 = Francis|first3 = D.S.|last3 = Kumar|arxiv = 1308.0090|s2cid = 7357110}}</ref><ref>{{Cite journal|title = Threshold Logic Computing: Memristive-CMOS Circuits for Fast Fourier Transform and Vedic Multiplication|journal = IEEE Transactions on Very Large Scale Integration (VLSI) Systems|date = 2015-11-01|issn = 1063-8210|pages = 2690–2694|volume = 23|issue = 11|doi = 10.1109/TVLSI.2014.2371857|first1 = A.P.|last1 = James|first2 = D.S.|last2 = Kumar|first3 = A.|last3 = Ajayan|arxiv = 1411.5255|s2cid = 6076956}}</ref><br />
<br />
There exist several neuron inspired threshold logic functions implemented with memristors that have applications in high level pattern recognition applications. Some of the applications reported recently include speech recognition, face recognition and object recognition. They also find applications in replacing conventional digital logic gates.<br />
<br />
利用记忆电阻器实现的神经元启发阈值逻辑函数在高级模式识别中有着广泛的应用。最近报道的一些应用包括语音识别、人脸识别和物体识别。它们还可以用来取代传统的数字逻辑门。<br />
<br />
For ideal passive memristive circuits there is an exact equation (Caravelli-Traversa-[[Di Ventra]] equation) for the internal memory of the circuit:<ref>{{cite journal |last=Caravelli |display-authors=etal|arxiv=1608.08651 |title=The complex dynamics of memristive circuits: analytical results and universal slow relaxation |year=2017 |doi=10.1103/PhysRevE.95.022140 |pmid= 28297937 |volume=95 |issue= 2 |pages= 022140 |journal=Physical Review E|bibcode=2017PhRvE..95b2140C |s2cid=6758362}}</ref><br />
<br />
For ideal passive memristive circuits there is an exact equation (Caravelli-Traversa-Di Ventra equation) for the internal memory of the circuit:<br />
<br />
对于理想的无源记忆电路，电路的内部记忆有一个精确的方程(Caravelli-Traversa-Di Ventra 方程) :<br />
<br />
:<math> \frac{d}{dt} \vec{W} = \alpha \vec{W}-\frac{1}{\beta} (I+\xi \Omega W)^{-1} \Omega \vec S </math><br />
<br />
:\frac{d}{dt} \vec{W} = \alpha \vec{W}-\frac{1}{\beta} (I+\xi \Omega W)^{-1} \Omega \vec S<br />
<br />
: \frac{d}{dt} \vec{W} = \alpha \vec{W}-\frac{1}{\beta} (I+\xi \Omega W)^{-1} \Omega \vec S<br />
<br />
as a function of the properties of the physical memristive network and the external sources. In the equation above, <math>\alpha</math> is the "forgetting" time scale constant, <math> \xi=r-1</math> and <math>r=\frac{R_\text{off}}{R_\text{on}}</math> is the ratio of ''off'' and ''on'' values of the limit resistances of the memristors, <math> \vec S </math> is the vector of the sources of the circuit and <math>\Omega</math> is a projector on the fundamental loops of the circuit. The constant <math>\beta</math> has the dimension of a voltage and is associated to the properties of the [[memristor]]; its physical origin is the charge mobility in the conductor. The diagonal matrix and vector <math>W=\operatorname{diag}(\vec W)</math> and <math>\vec W</math> respectively, are instead the internal value of the memristors, with values between 0 and 1. This equation thus requires adding extra constraints on the memory values in order to be reliable.<br />
<br />
<nowiki>as a function of the properties of the physical memristive network and the external sources. In the equation above, \alpha is the "forgetting" time scale constant, \xi=r-1 and r=\frac{R_\text{off}}{R_\text{on}} is the ratio of off and on values of the limit resistances of the memristors, \vec S is the vector of the sources of the circuit and \Omega is a projector on the fundamental loops of the circuit. The constant \beta has the dimension of a voltage and is associated to the properties of the memristor; its physical origin is the charge mobility in the conductor. The diagonal matrix and vector W=\operatorname{diag}(\vec W) and \vec W respectively, are instead the internal value of the memristors, with values between 0 and 1. This equation thus requires adding extra constraints on the memory values in order to be reliable.</nowiki><br />
<br />
<nowiki>作为物理记忆网络和外部源的性质的函数。在上述方程中，α 是“遗忘”时间尺度常数，xi = r-1，r = frac { r _ text { off }{ on }{ r _ text { on }}是记忆电阻器极限电阻的开关和开关值之比，vec s 是电路源的矢量，Omega 是电路基本环路的投影仪。常数 β 具有电压的尺寸，与记忆电阻器的特性有关; 它的物理起源是导体中的电荷迁移率。对角矩阵和向量 w = 操作者名{ diag }(vec w)和 vec w 分别是记忆电阻器的内值，值在0到1之间。因此，这个等式需要在内存值上添加额外的约束，以保证可靠性。</nowiki><br />
<br />
==See also ==<br />
{{Columns-list|colwidth=18em|<br />
* [[AI accelerator (computer hardware)|AI accelerator]]<br />
* [[Artificial brain]]<br />
* [[Biomorphic]]<br />
* [[Cognitive computer]]<br />
* [[Computation and Neural Systems]]<br />
* [[Differentiable programming]]<br />
* [[Event camera]]<br />
* [[Neurorobotics]]<br />
* [[Optical flow sensor]]<br />
* [[Physical neural network]]<br />
* [[SpiNNaker]]<br />
* [[SyNAPSE]]<br />
* [[Retinomorphic sensor]]<br />
* [[Vision chip]]<br />
* [[Vision processing unit]]<br />
* [[Zeroth (software)]]<br />
* [[Hardware for artificial intelligence]]<br />
}}<br />
<br />
{{Portal bar|Electronics}}<br />
<br />
==References==<br />
{{Reflist|40em}}<br />
<br />
==External links==<br />
<!--======================== {{No more links}} ============================<br />
| PLEASE BE CAUTIOUS IN ADDING MORE LINKS TO THIS ARTICLE. Wikipedia |<br />
| is not a collection of links nor should it be used for advertising. |<br />
| |<br />
| Excessive or inappropriate links WILL BE DELETED. |<br />
| See [[Wikipedia:External links]] & [[Wikipedia:Spam]] for details. |<br />
| |<br />
| If there are already plentiful links, please propose additions or |<br />
| replacements on this article's discussion page, or submit your link |<br />
| to the relevant category at the Open Directory Project (dmoz.org) |<br />
| and link back to that category using the {{dmoz}} template. |<br />
======================={{No more links}}=============================--><br />
{{Commons category|Neuromorphic Engineering}}<br />
*[https://web.archive.org/web/20150727034331/http://ine-web.org/workshops/workshops-overview Telluride Neuromorphic Engineering Workshop]<br />
*[https://archive.today/20130115190057/http://capocaccia.ethz.ch/ CapoCaccia Cognitive Neuromorphic Engineering Workshop]<br />
*[https://web.archive.org/web/20190716132350/http://www.ine-web.org/ Institute of Neuromorphic Engineering]<br />
*[http://www.ine-news.org/ INE news site].<br />
*[http://www.frontiersin.org/neuromorphic_engineering Frontiers in Neuromorphic Engineering Journal]<br />
*[http://www.cns.caltech.edu/ Computation and Neural Systems] department at the [[California Institute of Technology]].<br />
*[http://www.humanbrainproject.eu/ Human Brain Project official site]<br />
*[https://www.the-scientist.com/features/building-a-silicon-brain-65738 Building a Silicon Brain:] Computer chips based on biological neurons may help simulate larger and more-complex brain models. May 1, 2019. SANDEEP RAVINDRAN<br />
<br />
<br />
* Telluride Neuromorphic Engineering Workshop<br />
*CapoCaccia Cognitive Neuromorphic Engineering Workshop<br />
*Institute of Neuromorphic Engineering<br />
*INE news site.<br />
*Frontiers in Neuromorphic Engineering Journal<br />
*Computation and Neural Systems department at the California Institute of Technology.<br />
*Human Brain Project official site<br />
*Building a Silicon Brain: Computer chips based on biological neurons may help simulate larger and more-complex brain models. May 1, 2019. SANDEEP RAVINDRAN<br />
<br />
=<nowiki>= = 外部链接 =</nowiki>= <br />
*碲化物神经形态工程工作室<br />
*CapoCaccia 认知神经形态工程工作室<br />
*神经形态工程研究所<br />
*INE 新闻站点。<br />
*《神经形态工程学前沿》<br />
*加州理工学院计算与神经系统系。<br />
*人脑项目官方网站<br />
*打造硅脑: 基于生物神经元的计算机芯片可能有助于模拟更大、更复杂的大脑模型。2019年5月1日。SANDEEP RAVINDRAN<br />
<br />
{{Differentiable computing}}<br />
{{Authority control}}<br />
<br />
[[Category:Electrical engineering]]<br />
[[Category:Neuroscience]]<br />
[[分类:AI accelerators|*]]<br />
[[Category:Artificial intelligence]]<br />
[[Category:Robotics]]<br />
<br />
Category:Electrical engineering<br />
Category:Neuroscience<br />
*<br />
Category:Artificial intelligence<br />
Category:Robotics<br />
<br />
类别: 电气工程类别: 神经科学 * 类别: 人工智能类别: 机器人<br />
<br />
<noinclude><br />
<br />
<small>This page was moved from [[wikipedia:en:Neuromorphic engineering]]. Its edit history can be viewed at [[神经形态计算/edithistory]]</small></noinclude><br />
<br />
[[Category:待整理页面]]<br />
<br />
<references /></div>Shenky20https://wiki.swarma.org/index.php?title=%E7%A5%9E%E7%BB%8F%E5%BD%A2%E6%80%81%E8%AE%A1%E7%AE%97&diff=30172神经形态计算2022-04-13T07:09:28Z<p>Shenky20：</p>
<hr />
<div><br />
此词条由神经动力学模型读书会词条梳理志愿者Shenky20翻译审校，未经专家审校，带来阅读不便，请见谅。<br />
<br />
{{Use American English|date = January 2019}}<br />
{{Use mdy dates|date = January 2019}}<br />
<br />
'''Neuromorphic engineering''', also known as '''neuromorphic computing''',<ref>{{Cite journal | doi = 10.1145/2601069| title = Neuromorphic computing gets ready for the (really) big time| journal = [[Communications of the ACM]]| volume = 57| issue = 6| year = 2014| last1 = Monroe | first1 = D. | pages = 13–15| s2cid = 20051102}}</ref><ref>{{Cite journal | doi = 10.1088/0957-4484/21/17/175202| pmid = 20368686| title = Nanotube devices based crossbar architecture: Toward neuromorphic computing| journal = Nanotechnology| volume = 21| issue = 17| pages = 175202| year = 2010| last1 = Zhao | first1 = W. S. | last2 = Agnus | first2 = G. | last3 = Derycke | first3 = V. | last4 = Filoramo | first4 = A. | last5 = Bourgoin | first5 = J. -P. | last6 = Gamrat | first6 = C. | bibcode = 2010Nanot..21q5202Z| url = https://zenodo.org/record/3428659}}</ref><ref name="humanbrainproject">{{YouTube|id=6RoiZ90mGfw|title=The Human Brain Project SP 9: Neuromorphic Computing Platform}}</ref> is the use of [[very-large-scale integration]] (VLSI) systems containing electronic [[analog circuit]]s to mimic neuro-biological architectures present in the nervous system. A neuromorphic computer/chip is any device that uses physical artificial neurons (made from silicon) to do computations.<ref>{{cite journal|last1=Mead|first1=Carver|title=Neuromorphic electronic systems|journal=Proceedings of the IEEE|date=1990|volume=78|issue=10|pages=1629–1636|doi=10.1109/5.58356|url=https://authors.library.caltech.edu/53090/1/00058356.pdf}}</ref><ref name=":2" /> In recent times, the term ''neuromorphic'' has been used to describe analog, digital, mixed-mode analog/digital VLSI, and software systems that implement models of [[neural system]]s (for [[perception]], [[motor control]], or [[multisensory integration]]). The implementation of neuromorphic computing on the hardware level can be realized by oxide-based [[memristor]]s,<ref name="Maan 1–13">{{Cite journal|last1=Maan|first1=A. K.|last2=Jayadevi|first2=D. A.|last3=James|first3=A. P.|date=2016-01-01|title=A Survey of Memristive Threshold Logic Circuits|journal=IEEE Transactions on Neural Networks and Learning Systems|volume=PP|issue=99|pages=1734–1746|doi=10.1109/TNNLS.2016.2547842|pmid=27164608|issn=2162-237X|arxiv=1604.07121|bibcode=2016arXiv160407121M|s2cid=1798273}}</ref> spintronic memories, threshold switches, and [[transistor]]s.<ref>{{Cite journal|title = Mott Memory and Neuromorphic Devices|journal = Proceedings of the IEEE|date = 2015-08-01|issn = 0018-9219|pages = 1289–1310|volume = 103|issue = 8|doi = 10.1109/JPROC.2015.2431914|first1 = You|last1 = Zhou|first2 = S.|last2 = Ramanathan|s2cid = 11347598|url = https://zenodo.org/record/895565}}</ref><ref name=":2">{{Cite document|title=Neuromorphic Circuits With Neural Modulation Enhancing the Information Content of Neural Signaling {{!}} International Conference on Neuromorphic Systems 2020|language=EN|doi=10.1145/3407197.3407204|s2cid=220794387}}</ref> Training software-based neuromorphic systems of spiking neural networks can be achieved using error backpropagation, e.g., using [[Python (programming language)|Python]] based frameworks such as snnTorch,<ref>{{cite journal |last1=Eshraghian|first1=Jason K.|last2=Ward|first2=Max|last3=Neftci |first3=Emre|last4=Wang|first4=Xinxin|last5=Lenz|first5=Gregor|last6=Dwivedi|first6=Girish|last7=Bennamoun|first7=Mohammed|last8=Jeong|first8=Doo Seok|last9=Lu|first9=Wei D.|title=Training Spiking Neural Networks Using Lessons from Deep Learning |date=1 October 2021 |arxiv=2109.12894 }}</ref> or using canonical learning rules from the biological learning literature, e.g., using BindsNet.<ref>{{Cite web | url=https://github.com/Hananel-Hazan/bindsnet | title=Hananel-Hazan/bindsnet: Simulation of spiking neural networks (SNNs) using PyTorch.| date=31 March 2020}}</ref><br />
<br />
Neuromorphic engineering, also known as neuromorphic computing, is the use of very-large-scale integration (VLSI) systems containing electronic analog circuits to mimic neuro-biological architectures present in the nervous system. A neuromorphic computer/chip is any device that uses physical artificial neurons (made from silicon) to do computations. In recent times, the term neuromorphic has been used to describe analog, digital, mixed-mode analog/digital VLSI, and software systems that implement models of neural systems (for perception, motor control, or multisensory integration). The implementation of neuromorphic computing on the hardware level can be realized by oxide-based memristors, spintronic memories, threshold switches, and transistors. Training software-based neuromorphic systems of spiking neural networks can be achieved using error backpropagation, e.g., using Python based frameworks such as snnTorch, or using canonical learning rules from the biological learning literature, e.g., using BindsNet.<br />
<br />
'''<font color="#ff8000">神经形态工程Neuromorphic engineering</font>'''(也称为'''<font color="#ff8000">神经形态计算Neuromorphic computing</font>''')是指使用包含电子'''<font color="#ff8000">模拟电路Analog circuit</font>'''的'''<font color="#ff8000">超大规模集成电路Very-large-scale integration</font>'''系统来模拟神经系统中生理结构的研究方法。神经形态计算机或神经形态芯片包括任何使用由硅制成的人造神经元进行计算的设备。近年来，神经形态学（neuromorphic）这个术语被用来描述能够实现'''<font color="#ff8000">神经系统Neural system</font>'''模型功能(如'''<font color="#ff8000">感知Perception</font>'''、'''<font color="#ff8000">运动控制Motor control</font>'''，'''<font color="#ff8000">多感官整合Multisensory integration</font>'''等)的模拟、数字、模拟/数字混合模式超大规模集成电路和软件系统。神经形态计算的硬件实现可以通过基于氧化物的'''<font color="#ff8000">记忆电阻器Memristor</font>'''、自旋电子存储器、阈值开关和'''<font color="#ff8000">晶体管Transistor</font>'''来实现。对基于软件的脉冲神经网络系统的训练可以通过误差反向传播机制来实现，例如，使用snnTorch等基于Python的框架，或使用BindsNet等典型的受生物启发的学习模式。<br />
<br />
A key aspect of neuromorphic engineering is understanding how the morphology of individual neurons, circuits, applications, and overall architectures creates desirable computations, affects how information is represented, influences robustness to damage, incorporates learning and development, adapts to local change (plasticity), and facilitates evolutionary change.<br />
<br />
A key aspect of neuromorphic engineering is understanding how the morphology of individual neurons, circuits, applications, and overall architectures creates desirable computations, affects how information is represented, influences robustness to damage, incorporates learning and development, adapts to local change (plasticity), and facilitates evolutionary change.<br />
<br />
神经形态工程领域的一个关键问题，就是理解'''<font color="#32CD32">单个神经元形态、神经回路、应用和整体结构</font>'''如何产生理想的计算，如何影响信息的表示和对破坏的鲁棒性，如何整合学习和发展，如何适应局部变化(可塑性) 并促进逐渐发展的变化。<br />
<br />
Neuromorphic engineering is an interdisciplinary subject that takes inspiration from [[biology]], [[physics]], [[mathematics]], [[computer science]], and [[electronic engineering]]<ref name=":2" /> to design artificial neural systems, such as [[Machine vision|vision systems]], head-eye systems, auditory processors, and autonomous robots, whose physical architecture and design principles are based on those of biological nervous systems.<ref>{{Cite journal | doi = 10.1155/2012/705483| title = Qualitative Functional Decomposition Analysis of Evolved Neuromorphic Flight Controllers| journal = Applied Computational Intelligence and Soft Computing| volume = 2012| pages = 1–21| year = 2012| last1 = Boddhu | first1 = S. K. | last2 = Gallagher | first2 = J. C. | doi-access = free}}</ref> It was developed by [[Carver Mead]]<ref>{{cite web|title=carver mead website|last1=Mead|first1=Carver|url=http://www.carvermead.caltech.edu/index.html|website=carvermead}}</ref> in the late 1980s.<br />
<br />
Neuromorphic engineering is an interdisciplinary subject that takes inspiration from biology, physics, mathematics, computer science, and electronic engineering to design artificial neural systems, such as vision systems, head-eye systems, auditory processors, and autonomous robots, whose physical architecture and design principles are based on those of biological nervous systems. It was developed by Carver Mead in the late 1980s.<br />
<br />
神经形态工程是以生物学、物理学、数学、计算机科学和电子工程为基础，设计人工神经系统，如视觉系统、头眼系统、听觉处理器和自主机器人的一门交叉学科。它是由卡弗 · 米德在20世纪80年代后期开发的。<br />
<br />
== Neurological inspiration==<br />
Neuromorphic engineering is set apart by the inspiration it takes from what we know about the structure and operations of the [[brain]]. Neuromorphic engineering translates what we know about the brain's function into computer systems. Work has mostly focused on replicating the analog nature of [[biological computation]] and the role of [[neuron]]s in [[cognition]].<br />
<br />
Neuromorphic engineering is set apart by the inspiration it takes from what we know about the structure and operations of the brain. Neuromorphic engineering translates what we know about the brain's function into computer systems. Work has mostly focused on replicating the analog nature of biological computation and the role of neurons in cognition.<br />
<br />
神经学灵感工程学是从我们对大脑结构和运作的了解中获得灵感而形成的。神经形态工程学将我们对大脑功能的了解转化为计算机系统。工作主要集中在复制生物计算的模拟特性和神经元在认知中的作用。<br />
<br />
The biological processes of neurons and their [[synapse]]s are dauntingly complex, and thus very difficult to artificially simulate. A key feature of biological brains is that all of the processing in neurons use analog chemical signals. This makes it hard to replicate brains in computers because the current generation of computers is completely digital. However, the characteristics of these parts can be abstracted into mathematical functions that closely capture the essence of the neuron's operations.<br />
<br />
The biological processes of neurons and their synapses are dauntingly complex, and thus very difficult to artificially simulate. A key feature of biological brains is that all of the processing in neurons use analog chemical signals. This makes it hard to replicate brains in computers because the current generation of computers is completely digital. However, the characteristics of these parts can be abstracted into mathematical functions that closely capture the essence of the neuron's operations.<br />
<br />
神经元及其突触的生物学过程极其复杂，难以人工模拟。生物大脑的一个关键特征是，神经元的所有处理过程都使用模拟化学信号。这使得很难在计算机中复制大脑，因为目前的计算机是完全数字化的。然而，这些部分的特征可以抽象为数学函数，这些函数紧密地抓住了神经元运行的本质。<br />
<br />
The goal of neuromorphic computing is not to perfectly mimic the brain and all of its functions, but instead to extract what is known of its structure and operations to be used in a practical computing system. No neuromorphic system will claim nor attempt to reproduce every element of neurons and synapses, but all adhere to the idea that computation is highly [[distributed processing|distributed]] throughout a series of small computing elements analogous to a neuron. While this sentiment is standard, researchers chase this goal with different methods.<ref>{{Cite journal | doi = 10.1088/1741-2560/13/5/051001| title = Large-scale neuromorphic computing systems| journal = Journal of Neural Engineering| volume = 13| pages = 1–15| year = 2016| last1 = Furber | first1 = Steve| issue = 5| pmid = 27529195| bibcode = 2016JNEng..13e1001F| doi-access = free}}</ref><br />
<br />
The goal of neuromorphic computing is not to perfectly mimic the brain and all of its functions, but instead to extract what is known of its structure and operations to be used in a practical computing system. No neuromorphic system will claim nor attempt to reproduce every element of neurons and synapses, but all adhere to the idea that computation is highly distributed throughout a series of small computing elements analogous to a neuron. While this sentiment is standard, researchers chase this goal with different methods.<br />
<br />
神经形态计算的目标不是完美地模拟大脑及其所有功能，而是提取已知的大脑结构和操作，用于实际的计算系统。没有哪个神经形态学系统会声称或试图复制神经元和突触的每一个元素，但所有人都坚持这样的观点，即计算是高度分布在一系列类似于神经元的小型计算元素中的。虽然这种情绪是标准的，但研究人员用不同的方法追求这一目标。<br />
<br />
==Examples==<br />
As early as 2006, researchers at [[Georgia Tech]] published a field programmable neural array.<ref>{{Cite book|title = A field programmable neural array|last1 = Farquhar|first1 = Ethan|date = May 2006|journal = IEEE International Symposium on Circuits and Systems|pages = 4114–4117|last2 = Hasler|first2 = Paul.|doi = 10.1109/ISCAS.2006.1693534|isbn = 978-0-7803-9389-9|s2cid = 206966013}}</ref> This chip was the first in a line of increasingly complex arrays of floating gate transistors that allowed programmability of charge on the gates of [[MOSFET]]s to model the channel-ion characteristics of neurons in the brain and was one of the first cases of a silicon programmable array of neurons.<br />
<br />
As early as 2006, researchers at Georgia Tech published a field programmable neural array. This chip was the first in a line of increasingly complex arrays of floating gate transistors that allowed programmability of charge on the gates of MOSFETs to model the channel-ion characteristics of neurons in the brain and was one of the first cases of a silicon programmable array of neurons.<br />
<br />
早在2006年，佐治亚理工学院的研究人员就发表了现场可编程神经阵列。这种芯片是一系列越来越复杂的浮栅晶体管阵列中的第一个，这些晶体管可以在 mosfet 的栅极上编程来模拟大脑中神经元的通道离子特性，这也是硅可编程神经元阵列的首批案例之一。<br />
<br />
In November 2011, a group of [[MIT]] researchers created a computer chip that mimics the analog, ion-based communication in a synapse between two neurons using 400 transistors and standard [[CMOS]] manufacturing techniques.<ref>{{cite web|title=MIT creates "brain chip"|url=http://www.extremetech.com/extreme/105067-mit-creates-brain-chip|access-date=4 December 2012}}</ref><ref name="Neuromorphic silicon paper">{{cite journal|title=Neuromorphic silicon neurons and large-scale neural networks: challenges and opportunities|doi=10.3389/fnins.2011.00108|pmid=21991244|pmc=3181466|volume=5|pages=108|journal=Frontiers in Neuroscience|year=2011|last1=Poon|first1=Chi-Sang|last2=Zhou|first2=Kuan|doi-access=free}}</ref><br />
<br />
In November 2011, a group of MIT researchers created a computer chip that mimics the analog, ion-based communication in a synapse between two neurons using 400 transistors and standard CMOS manufacturing techniques.<br />
<br />
2011年11月，麻省理工学院的一组研究人员发明了一种计算机芯片，该芯片使用400个晶体管和标准的 CMOS 制造技术，在两个神经元之间的突触中模拟模拟基于离子的通讯。<br />
<br />
In June 2012, [[spintronic]] researchers at [[Purdue University]] presented a paper on the design of a neuromorphic chip using [[Spin valve|lateral spin valve]]s and [[memristor]]s. They argue that the architecture works similarly to neurons and can therefore be used to test methods of reproducing the brain's processing. In addition, these chips are significantly more energy-efficient than conventional ones.<ref name="Spin Devices Prop">{{Cite arXiv|title=Proposal For Neuromorphic Hardware Using Spin Devices|eprint=1206.3227|last1=Sharad|first1=Mrigank|last2=Augustine|first2=Charles|last3=Panagopoulos|first3=Georgios|last4=Roy|first4=Kaushik|class=cond-mat.dis-nn|year=2012}}</ref><br />
<br />
In June 2012, spintronic researchers at Purdue University presented a paper on the design of a neuromorphic chip using lateral spin valves and memristors. They argue that the architecture works similarly to neurons and can therefore be used to test methods of reproducing the brain's processing. In addition, these chips are significantly more energy-efficient than conventional ones.<br />
<br />
2012年6月，普渡大学的自旋电子学研究人员发表了一篇关于利用侧向自旋阀和记忆电阻器设计神经形态芯片的论文。他们认为，这种结构的工作原理与神经元相似，因此可以用来测试复制大脑处理过程的方法。此外，这些芯片明显比传统芯片更节能。<br />
<br />
Research at [[HP Labs]] on Mott memristors has shown that while they can be non-[[Volatile memory|volatile]], the volatile behavior exhibited at temperatures significantly below the [[phase transition]] temperature can be exploited to fabricate a [[neuristor]],<ref name=":0" /> a biologically-inspired device that mimics behavior found in neurons.<ref name=":0">{{Cite journal | doi = 10.1038/nmat3510| pmid = 23241533| title = A scalable neuristor built with Mott memristors| journal = Nature Materials| volume = 12| issue = 2| pages = 114–7| year = 2012| last1 = Pickett | first1 = M. D. | last2 = Medeiros-Ribeiro | first2 = G. | last3 = Williams | first3 = R. S. | bibcode = 2013NatMa..12..114P| s2cid = 16271627| url = https://semanticscholar.org/paper/b6ba6f496ace2b947f111059663e76bb60e9efeb}}</ref> In September 2013, they presented models and simulations that show how the spiking behavior of these neuristors can be used to form the components required for a [[Turing machine]].<ref>{{cite journal|doi=10.1088/0957-4484/24/38/384002|title=Phase transitions enable computational universality in neuristor-based cellular automata|author1=Matthew D Pickett|author2=R Stanley Williams|name-list-style=amp|date=September 2013|publisher=IOP Publishing Ltd|journal=Nanotechnology|volume=24|issue=38|pmid=23999059|bibcode=2013Nanot..24L4002P|at=384002}}</ref><br />
<br />
Research at HP Labs on Mott memristors has shown that while they can be non-volatile, the volatile behavior exhibited at temperatures significantly below the phase transition temperature can be exploited to fabricate a neuristor, a biologically-inspired device that mimics behavior found in neurons. In September 2013, they presented models and simulations that show how the spiking behavior of these neuristors can be used to form the components required for a Turing machine.<br />
<br />
惠普实验室在莫特记忆电阻器上的研究表明，尽管它们可以是非挥发性的，但是在相变温度以下显示的挥发性行为可以被用来制造神经元电阻器，这是一种模仿神经元行为的生物设备。2013年9月，他们展示了模型和仿真，展示了这些神经元的尖峰行为是如何被用来形成图灵机所需的元件的。<br />
<br />
[[Neurogrid]], built by ''Brains in Silicon'' at [[Stanford University]],<ref>{{cite journal|last1=Boahen|first1=Kwabena|title=Neurogrid: A Mixed-Analog-Digital Multichip System for Large-Scale Neural Simulations|journal=Proceedings of the IEEE|date=24 April 2014|volume=102|issue=5|pages=699–716|doi=10.1109/JPROC.2014.2313565|s2cid=17176371}}</ref> is an example of hardware designed using neuromorphic engineering principles. The circuit board is composed of 16 custom-designed chips, referred to as NeuroCores. Each NeuroCore's analog circuitry is designed to emulate neural elements for 65536 neurons, maximizing energy efficiency. The emulated neurons are connected using digital circuitry designed to maximize spiking throughput.<ref>{{cite journal|doi=10.1038/503022a|pmid = 24201264|title = Neuroelectronics: Smart connections|journal = Nature|volume = 503|issue = 7474|pages = 22–4|year = 2013|last1 = Waldrop|first1 = M. Mitchell|bibcode = 2013Natur.503...22W|doi-access = free}}</ref><ref>{{cite journal|doi=10.1109/JPROC.2014.2313565|title = Neurogrid: A Mixed-Analog-Digital Multichip System for Large-Scale Neural Simulations|journal = Proceedings of the IEEE|volume = 102|issue = 5|pages = 699–716|year = 2014|last1 = Benjamin|first1 = Ben Varkey|last2 = Peiran Gao|last3 = McQuinn|first3 = Emmett|last4 = Choudhary|first4 = Swadesh|last5 = Chandrasekaran|first5 = Anand R.|last6 = Bussat|first6 = Jean-Marie|last7 = Alvarez-Icaza|first7 = Rodrigo|last8 = Arthur|first8 = John V.|last9 = Merolla|first9 = Paul A.|last10 = Boahen|first10 = Kwabena|s2cid = 17176371}}</ref><br />
<br />
Neurogrid, built by Brains in Silicon at Stanford University, is an example of hardware designed using neuromorphic engineering principles. The circuit board is composed of 16 custom-designed chips, referred to as NeuroCores. Each NeuroCore's analog circuitry is designed to emulate neural elements for 65536 neurons, maximizing energy efficiency. The emulated neurons are connected using digital circuitry designed to maximize spiking throughput.<br />
<br />
神经网格是由斯坦福大学硅谷大脑公司建立的，是一个使用神经形态工程原理设计硬件的例子。该电路板由16个定制设计的芯片组成，称为 NeuroCores。每个 NeuroCore 的模拟电路被设计为模拟65536个神经元的神经元元件，最大限度地提高能量效率。模拟的神经元通过设计的数字电路连接，以最大化脉冲吞吐量。<br />
<br />
A research project with implications for neuromorphic engineering is the [[Human Brain Project]] that is attempting to simulate a complete human brain in a supercomputer using biological data. It is made up of a group of researchers in neuroscience, medicine, and computing.<ref>{{cite web|title=Involved Organizations|url=http://www.humanbrainproject.eu/partners.html|access-date=22 February 2013|url-status=dead|archive-url=https://web.archive.org/web/20130302142627/http://www.humanbrainproject.eu/partners.html|archive-date=2 March 2013}}</ref> [[Henry Markram]], the project's co-director, has stated that the project proposes to establish a foundation to explore and understand the brain and its diseases, and to use that knowledge to build new computing technologies. The three primary goals of the project are to better understand how the pieces of the brain fit and work together, to understand how to objectively diagnose and treat brain diseases, and to use the understanding of the human brain to develop neuromorphic computers. That the simulation of a complete human brain will require a supercomputer a thousand times more powerful than today's encourages the current focus on neuromorphic computers.<ref>{{cite web|title=Human Brain Project|url=http://www.humanbrainproject.eu|access-date=22 February 2013}}</ref> $1.3 billion has been allocated to the project by The [[European Commission]].<ref>{{cite web|title=The Human Brain Project and Recruiting More Cyberwarriors|url=http://www.marketplace.org/topics/tech/human-brain-project-and-recruiting-more-cyberwarriors|access-date=22 February 2013|date=January 29, 2013}}</ref><br />
<br />
A research project with implications for neuromorphic engineering is the Human Brain Project that is attempting to simulate a complete human brain in a supercomputer using biological data. It is made up of a group of researchers in neuroscience, medicine, and computing. Henry Markram, the project's co-director, has stated that the project proposes to establish a foundation to explore and understand the brain and its diseases, and to use that knowledge to build new computing technologies. The three primary goals of the project are to better understand how the pieces of the brain fit and work together, to understand how to objectively diagnose and treat brain diseases, and to use the understanding of the human brain to develop neuromorphic computers. That the simulation of a complete human brain will require a supercomputer a thousand times more powerful than today's encourages the current focus on neuromorphic computers. $1.3 billion has been allocated to the project by The European Commission.<br />
<br />
一个对神经形态工程有影响的研究项目是人脑项目，它试图用生物数据在超级计算机中模拟完整的人脑。它由一群神经科学、医学和计算机科学的研究人员组成。该项目的联合主管亨利•马克拉姆(Henry Markram)表示，该项目提议建立一个基础，以探索和了解大脑及其疾病，并利用这些知识构建新的计算机技术。这个项目的三个主要目标是: 更好地理解大脑的各个部分是如何相互匹配和协同工作的; 理解如何客观地诊断和治疗脑部疾病; 以及利用对人类大脑的理解来开发神经形态计算机。模拟一个完整的人类大脑需要一台比现在强大一千倍的超级计算机，这鼓励了当前对神经形态计算机的关注。欧洲委员会已经拨款13亿美元用于这个项目。<br />
<br />
Other research with implications for neuromorphic engineering involves the [[BRAIN Initiative]]<ref name="economist">[https://www.economist.com/news/science-and-technology/21582495-computers-will-help-people-understand-brains-better-and-understanding-brains Neuromorphic computing: The machine of a new soul], The Economist, 2013-08-03</ref> and the [[TrueNorth]] chip from [[IBM]].<ref>{{cite journal|last1=Modha|first1=Dharmendra|title=A million spiking-neuron integrated circuit with a scalable communication network and interface|journal=Science|date=Aug 2014|volume=345|issue=6197|pages=668–673|doi=10.1126/science.1254642|pmid=25104385|bibcode=2014Sci...345..668M|s2cid=12706847}}</ref> Neuromorphic devices have also been demonstrated using nanocrystals, nanowires, and conducting polymers.<ref>{{Cite web|url=http://jessamynfairfield.com/wp-content/uploads/2017/03/PWMar17Fairfield.pdf|title=Smarter Machines|last=Fairfield|first=Jessamyn|date=March 1, 2017}}</ref><br />
<br />
Other research with implications for neuromorphic engineering involves the BRAIN InitiativeNeuromorphic computing: The machine of a new soul, The Economist, 2013-08-03 and the TrueNorth chip from IBM. Neuromorphic devices have also been demonstrated using nanocrystals, nanowires, and conducting polymers.<br />
<br />
其他与神经形态工程有关的研究还包括 BRAIN initiativity/euromorphic computing: The machine of a new soul，The Economist，2013-08-03 and The TrueNorth chip from IBM。神经形态学设备也已经被证明使用纳米晶体、纳米线和导电聚合物。<br />
<br />
[[Intel]] unveiled its neuromorphic research chip, called “[[Intel Loihi|Loihi]]”, in October 2017. The chip uses an asynchronous [[spiking neural network]] (SNN) to implement adaptive self-modifying event-driven fine-grained parallel computations used to implement learning and inference with high efficiency.<ref>{{cite journal |last1=Davies |first1=Mike |title=Loihi: A Neuromorphic Manycore Processor with On-Chip Learning |journal=IEEE Micro |date=January 16, 2018 |volume=38 |issue=1 |pages=82–99 |display-authors=etal|doi=10.1109/MM.2018.112130359 |s2cid=3608458 }}</ref><ref name="Morris2017">{{cite web |last1=Morris |first1=John |title=Why Intel built a neuromorphic chip |url=https://www.zdnet.com/article/why-intel-built-a-neuromorphic-chip/ |website=ZDNet |access-date=17 August 2018 |language=en}}</ref><br />
<br />
Intel unveiled its neuromorphic research chip, called “Loihi”, in October 2017. The chip uses an asynchronous spiking neural network (SNN) to implement adaptive self-modifying event-driven fine-grained parallel computations used to implement learning and inference with high efficiency.<br />
<br />
2017年10月，英特尔公布了其神经形态研究芯片“ Loihi”。该芯片采用异步脉冲神经网络(SNN)实现自适应自修改事件驱动的细粒度并行计算，实现了高效的学习和推理。<br />
<br />
[[IMEC]], a Belgium-based nanoelectronics research center, demonstrated the world's first self-learning neuromorphic chip. The brain-inspired chip, based on OxRAM technology, has the capability of self-learning and has been demonstrated to have the ability to compose music.<ref>{{cite web |title=Imec demonstrates self-learning neuromorphic chip that composes music |url=https://www.imec-int.com/en/articles/imec-demonstrates-self-learning-neuromorphic-chip-that-composes-music |website=IMEC International |access-date=1 October 2019}}</ref> IMEC released the 30-second tune composed by the prototype. The chip was sequentially loaded with songs in the same time signature and style. The songs were old Belgian and French flute minuets, from which the chip learned the rules at play and then applied them.<ref>{{cite web|last1=Bourzac|first1=Katherine|title=A Neuromorphic Chip That Makes Music|url=https://spectrum.ieee.org/a-neuromorphic-chip-that-makes-music|url-status=live|access-date=1 October 2019|website=IEEE Spectrum|date=May 23, 2017}}</ref><br />
<br />
IMEC, a Belgium-based nanoelectronics research center, demonstrated the world's first self-learning neuromorphic chip. The brain-inspired chip, based on OxRAM technology, has the capability of self-learning and has been demonstrated to have the ability to compose music. IMEC released the 30-second tune composed by the prototype. The chip was sequentially loaded with songs in the same time signature and style. The songs were old Belgian and French flute minuets, from which the chip learned the rules at play and then applied them.<br />
<br />
比利时的一个纳米电子学研究中心 IMEC 展示了世界上第一个自学习神经形态芯片。这种基于 OxRAM 技术的大脑启发芯片具有自学习能力，并已被证明具有创作音乐的能力。IMEC 发布了由原型机谱写的30秒曲调。芯片按顺序加载同时签名和风格的歌曲。歌曲是古老的比利时和法国长笛小步舞曲，筹码从中学习游戏规则，然后应用它们。<br />
<br />
[[Blue Brain Project|The Blue Brain Project]], led by Henry Markram, aims to build biologically detailed digital reconstructions and simulations of the mouse brain. The Blue Brain Project has created in silico models of rodent brains, while attempting to replicate as many details about its biology as possible. The supercomputer-based simulations offer new perspectives on understanding the structure and functions of the brain.<br />
<br />
The Blue Brain Project, led by Henry Markram, aims to build biologically detailed digital reconstructions and simulations of the mouse brain. The Blue Brain Project has created in silico models of rodent brains, while attempting to replicate as many details about its biology as possible. The supercomputer-based simulations offer new perspectives on understanding the structure and functions of the brain.<br />
<br />
由亨利 · 马克拉姆领导的蓝脑计划旨在建立小鼠大脑的生物细节数字重建和模拟。蓝色大脑项目已经在啮齿动物大脑的电子模型中创建，同时试图尽可能多地复制其生物学细节。基于超级计算机的模拟为理解大脑的结构和功能提供了新的视角。<br />
<br />
The European Union funded a series of projects at the University of Heidelberg, which led to the development of <br />
[[BrainScaleS]] (brain-inspired multiscale computation in neuromorphic hybrid systems), a hybrid analog [[neuromorphic]] supercomputer located at Heidelberg University, Germany. It was developed as part of the [[Human Brain Project]] neuromorphic computing platform and is the complement to the [[SpiNNaker]] supercomputer (which is based on digital technology). The architecture used in BrainScaleS mimics biological neurons and their connections on a physical level; additionally, since the components are made of silicon, these model neurons operate on average 864 times (24 hours of real time is 100 seconds in the machine simulation) that of their biological counterparts.<ref>{{Cite web|date=2016-03-21|title=Beyond von Neumann, Neuromorphic Computing Steadily Advances|url=https://www.hpcwire.com/2016/03/21/lacking-breakthrough-neuromorphic-computing-steadily-advance/|access-date=2021-10-08|website=HPCwire|language=en-US}}</ref><br />
<br />
The European Union funded a series of projects at the University of Heidelberg, which led to the development of <br />
BrainScaleS (brain-inspired multiscale computation in neuromorphic hybrid systems), a hybrid analog neuromorphic supercomputer located at Heidelberg University, Germany. It was developed as part of the Human Brain Project neuromorphic computing platform and is the complement to the SpiNNaker supercomputer (which is based on digital technology). The architecture used in BrainScaleS mimics biological neurons and their connections on a physical level; additionally, since the components are made of silicon, these model neurons operate on average 864 times (24 hours of real time is 100 seconds in the machine simulation) that of their biological counterparts.<br />
<br />
欧盟资助了海德堡大学的一系列项目，这些项目导致了 BrainScaleS (神经形态混合系统中受大脑启发的多尺度计算)的发展，这是一台位于德国海德堡大学的混合模拟神经形态超级计算机。它是作为人脑计划神经形态计算平台的一部分而开发的，是 SpiNNaker 超级计算机(基于数字技术)的补充。大脑尺度中使用的体系结构模拟了生物神经元及其在物理层面上的连接; 此外，由于这些组件是由硅制成的，这些模型神经元平均运行864倍(在机器模拟中，24小时的实时时间是100秒)。<br />
<br />
===Neuromorphic sensors===<br />
The concept of neuromorphic systems can be extended to sensors (not just to computation). An example of this applied to detecting [[light]] is the [[retinomorphic sensor]] or, when employed in an array, the [[event camera]].<br />
<br />
The concept of neuromorphic systems can be extended to sensors (not just to computation). An example of this applied to detecting light is the retinomorphic sensor or, when employed in an array, the event camera.<br />
<br />
= = = 神经形态传感器 = = = 神经形态系统的概念可以扩展到传感器(而不仅仅是计算)。用于检测光线的一个例子是视网膜变形传感器，或者在阵列中使用的事件摄像机。<br />
<br />
==Ethical considerations==<br />
While the interdisciplinary concept of neuromorphic engineering is relatively new, many of the same ethical considerations apply to neuromorphic systems as apply to [[human-like machines]] and [[artificial intelligence]] in general. However, the fact that neuromorphic systems are designed to mimic a [[human brain]] gives rise to unique ethical questions surrounding their usage.<br />
<br />
While the interdisciplinary concept of neuromorphic engineering is relatively new, many of the same ethical considerations apply to neuromorphic systems as apply to human-like machines and artificial intelligence in general. However, the fact that neuromorphic systems are designed to mimic a human brain gives rise to unique ethical questions surrounding their usage.<br />
<br />
虽然神经形态工程的跨学科概念相对较新，但许多同样的伦理考虑适用于神经形态系统，就像适用于类人机器和一般人工智能一样。然而，神经形态系统是为了模仿人类大脑而设计的这一事实引起了围绕其使用的独特的伦理问题。<br />
<br />
However, the practical debate is that neuromorphic hardware as well as artificial "neural networks" are immensely simplified models of how the brain operates or processes information at a much lower [[complex system|complexity in terms of size and functional technology]] and a much more regular structure in terms of [[brain connectivity|connectivity]]. Comparing [[neuromorphic chip]]s to the brain is a very crude comparison similar to comparing a plane to a bird just because they both have wings and a tail. The fact is that neural cognitive systems are many orders of magnitude more [[Energy efficiency (physics)|energy-]] and compute-efficient than current state-of-the-art AI and neuromorphic engineering is an attempt to narrow this gap by inspiring from the brain's mechanism just like many engineering designs have [[bioengineering|bio-inspired features]].<br />
<br />
However, the practical debate is that neuromorphic hardware as well as artificial "neural networks" are immensely simplified models of how the brain operates or processes information at a much lower complexity in terms of size and functional technology and a much more regular structure in terms of connectivity. Comparing neuromorphic chips to the brain is a very crude comparison similar to comparing a plane to a bird just because they both have wings and a tail. The fact is that neural cognitive systems are many orders of magnitude more energy- and compute-efficient than current state-of-the-art AI and neuromorphic engineering is an attempt to narrow this gap by inspiring from the brain's mechanism just like many engineering designs have bio-inspired features.<br />
<br />
然而，实际上的争论是，神经形态硬件和人工“神经网络”是大脑如何运作或处理信息的极其简化的模型，在大小和功能技术方面的复杂性要低得多，在连接方面的结构则更加规则。将神经形态芯片与大脑进行比较是一种非常粗糙的比较，类似于仅仅因为一架飞机有翅膀和一条尾巴就将它与一只鸟进行比较。事实上，神经认知系统比当前最先进的人工智能具有更多的能量和计算效率，而神经形态工程是一种通过从大脑机制中激发灵感来缩小这种差距的尝试，就像许多工程设计具有生物启发的特征一样。数量级。<br />
<br />
===Democratic concerns===<br />
Significant ethical limitations may be placed on neuromorphic engineering due to public perception.<ref>{{Cite report|url=https://ai100.stanford.edu/sites/g/files/sbiybj9861/f/ai_100_report_0831fnl.pdf|title=Artificial Intelligence and Life in 2030|author=2015 Study Panel|date=September 2016|work=One Hundred Year Study on Artificial Intelligence (AI100)|publisher=Stanford University}}</ref> Special [[Eurobarometer]] 382: Public Attitudes Towards Robots, a survey conducted by the European Commission, found that 60% of [[European Union]] citizens wanted a ban of robots in the care of children, the elderly, or the disabled. Furthermore, 34% were in favor of a ban on robots in education, 27% in healthcare, and 20% in leisure. The European Commission classifies these areas as notably “human.” The report cites increased public concern with robots that are able to mimic or replicate human functions. Neuromorphic engineering, by definition, is designed to replicate the function of the human brain.<ref name=":1">{{Cite web|url=http://ec.europa.eu/commfrontoffice/publicopinion/archives/ebs/ebs_382_en.pdf|title=Special Eurobarometer 382: Public Attitudes Towards Robots|last=European Commission|date=September 2012|website=European Commission}}</ref><br />
<br />
Significant ethical limitations may be placed on neuromorphic engineering due to public perception. Special Eurobarometer 382: Public Attitudes Towards Robots, a survey conducted by the European Commission, found that 60% of European Union citizens wanted a ban of robots in the care of children, the elderly, or the disabled. Furthermore, 34% were in favor of a ban on robots in education, 27% in healthcare, and 20% in leisure. The European Commission classifies these areas as notably “human.” The report cites increased public concern with robots that are able to mimic or replicate human functions. Neuromorphic engineering, by definition, is designed to replicate the function of the human brain.<br />
<br />
由于公众的认知，神经形态工程学可能会受到严重的伦理限制。欧盟委员会进行的一项调查发现，60% 的欧盟公民希望禁止照顾儿童、老人或残疾人的机器人。此外，34% 的人支持禁止机器人用于教育，27% 的人支持医疗保健，20% 的人支持休闲。欧盟委员会将这些地区明显归类为“人类”报告指出，公众越来越关注能够模仿或复制人类功能的机器人。神经形态工程，顾名思义，是为了复制人脑的功能而设计的。<br />
<br />
The democratic concerns surrounding neuromorphic engineering are likely to become even more profound in the future. The European Commission found that EU citizens between the ages of 15 and 24 are more likely to think of robots as human-like (as opposed to instrument-like) than EU citizens over the age of 55. When presented an image of a robot that had been defined as human-like, 75% of EU citizens aged 15–24 said it corresponded with the idea they had of robots while only 57% of EU citizens over the age of 55 responded the same way. The human-like nature of neuromorphic systems, therefore, could place them in the categories of robots many EU citizens would like to see banned in the future.<ref name=":1" /><br />
<br />
The democratic concerns surrounding neuromorphic engineering are likely to become even more profound in the future. The European Commission found that EU citizens between the ages of 15 and 24 are more likely to think of robots as human-like (as opposed to instrument-like) than EU citizens over the age of 55. When presented an image of a robot that had been defined as human-like, 75% of EU citizens aged 15–24 said it corresponded with the idea they had of robots while only 57% of EU citizens over the age of 55 responded the same way. The human-like nature of neuromorphic systems, therefore, could place them in the categories of robots many EU citizens would like to see banned in the future.<br />
<br />
围绕神经形态工程的民主关注可能在未来变得更加深刻。欧盟委员会(European Commission)发现，15至24岁的欧盟公民比55岁以上的欧盟公民更有可能认为机器人像人(而不是像仪器)。当看到一张被定义为“类人”的机器人图片时，年龄在15岁至24岁之间的欧盟公民中有75% 的人表示，这与他们对机器人的想法相符，而55岁以上的欧盟公民中只有57% 的人有同样的反应。因此，类似人类的神经形态系统，可以把它们归入许多欧盟公民希望在未来禁止使用的机器人类别。<br />
<br />
===Personhood ===<br />
As neuromorphic systems have become increasingly advanced, some scholars{{who|date=August 2021}} have advocated for granting [[personhood]] rights to these systems. If the brain is what grants humans their personhood, to what extent does a neuromorphic system have to mimic the human brain to be granted personhood rights? Critics of technology development in the [[Human Brain Project]], which aims to advance brain-inspired computing, have argued that advancement in neuromorphic computing could lead to machine consciousness or personhood.<ref>{{Cite journal|last=Aicardi|first=Christine|date=September 2018|title=Accompanying technology development in the Human Brain Project: From foresight to ethics management|journal=Futures|volume=102|pages=114–124|doi=10.1016/j.futures.2018.01.005|doi-access=free}}</ref> If these systems are to be treated as people, critics argue, then many tasks humans perform using neuromorphic systems, including the act of termination of neuromorphic systems, may be morally impermissible as these acts would violate the autonomy of the neuromorphic systems.<ref>{{Cite journal|last=Lim|first=Daniel|date=2014-06-01|title=Brain simulation and personhood: a concern with the Human Brain Project|journal=Ethics and Information Technology|language=en|volume=16|issue=2|pages=77–89|doi=10.1007/s10676-013-9330-5|s2cid=17415814|issn=1572-8439}}</ref><br />
<br />
As neuromorphic systems have become increasingly advanced, some scholars have advocated for granting personhood rights to these systems. If the brain is what grants humans their personhood, to what extent does a neuromorphic system have to mimic the human brain to be granted personhood rights? Critics of technology development in the Human Brain Project, which aims to advance brain-inspired computing, have argued that advancement in neuromorphic computing could lead to machine consciousness or personhood. If these systems are to be treated as people, critics argue, then many tasks humans perform using neuromorphic systems, including the act of termination of neuromorphic systems, may be morally impermissible as these acts would violate the autonomy of the neuromorphic systems.<br />
<br />
= = = 人格 = = 随着神经形态系统的日益发展，一些学者主张赋予这些系统人格权。如果是大脑赋予了人类人格，那么神经形态系统在多大程度上必须模仿人类大脑才能被赋予人格权利？“人脑计划”旨在推进以大脑为灵感的计算机技术发展，该计划的批评者认为，神经形态计算机技术的进步可能导致机器意识或人格的形成。批评家认为，如果这些系统被当作人来对待，那么人类使用神经形态系统执行的许多任务，包括终止神经形态系统的行为，可能在道德上是不允许的，因为这些行为将违反神经形态系统的自主性。<br />
<br />
==Dual use (military applications)==<br />
The [[Joint Artificial Intelligence Center]], a branch of the U.S. military, is a center dedicated to the procurement and implementation of AI software and neuromorphic hardware for combat use. Specific applications include smart headsets/goggles and robots. JAIC intends to rely heavily on neuromorphic technology to connect "every fighter every shooter" within a network of neuromorphic-enabled units.<br />
<br />
The Joint Artificial Intelligence Center, a branch of the U.S. military, is a center dedicated to the procurement and implementation of AI software and neuromorphic hardware for combat use. Specific applications include smart headsets/goggles and robots. JAIC intends to rely heavily on neuromorphic technology to connect "every fighter every shooter" within a network of neuromorphic-enabled units.<br />
<br />
军民两用联合人工智能中心是美国军队的一个分支，专门从事采购和实施用于战斗的人工智能软件和神经形态硬件。具体应用包括智能耳机/护目镜和机器人。JAIC 打算严重依赖神经形态技术来连接神经形态单位网络中的“每个战士每个射手”。<br />
<br />
== Legal considerations==<br />
Skeptics have argued that there is no way to apply the electronic personhood, the concept of personhood that would apply to neuromorphic technology, legally. In a letter signed by 285 experts in law, robotics, medicine, and ethics opposing a European Commission proposal to recognize “smart robots” as legal persons, the authors write, “A legal status for a robot can’t derive from the [[Natural person|Natural Person]] model, since the robot would then hold [[human rights]], such as the right to dignity, the right to its integrity, the right to remuneration or the right to citizenship, thus directly confronting the Human rights. This would be in contradiction with the [[Charter of Fundamental Rights of the European Union]] and the [[Convention for the Protection of Human Rights and Fundamental Freedoms]].”<ref>{{Cite web|url=http://www.robotics-openletter.eu/|title=Robotics Openletter {{!}} Open letter to the European Commission|language=fr-FR|access-date=2019-05-10}}</ref><br />
<br />
Skeptics have argued that there is no way to apply the electronic personhood, the concept of personhood that would apply to neuromorphic technology, legally. In a letter signed by 285 experts in law, robotics, medicine, and ethics opposing a European Commission proposal to recognize “smart robots” as legal persons, the authors write, “A legal status for a robot can’t derive from the Natural Person model, since the robot would then hold human rights, such as the right to dignity, the right to its integrity, the right to remuneration or the right to citizenship, thus directly confronting the Human rights. This would be in contradiction with the Charter of Fundamental Rights of the European Union and the Convention for the Protection of Human Rights and Fundamental Freedoms.”<br />
<br />
= = 法律方面的考虑 = = 怀疑论者认为，在法律上没有办法应用电子人格，这个人格概念将适用于神经形态技术。在一封由285名法律、机器人、医学和伦理学专家签名的信中，作者们反对欧盟委员会承认“智能机器人”为法人的提议，他们写道，“机器人的法律地位不能从自然人模型中推导出来，因为机器人将拥有人权，如尊严权、完整权、报酬权或公民权，从而直接面对人权。这将有悖于《欧洲联盟基本权利宪章和《保护人权和基本自由公约》。”<br />
<br />
===Ownership and property rights===<br />
There is significant legal debate around property rights and artificial intelligence. In ''Acohs Pty Ltd v. Ucorp Pty Ltd'', Justice Christopher Jessup of the [[Federal Court of Australia]] found that the [[source code]] for [[Material safety data sheets|Material Safety Data Sheets]] could not be [[Copyright law of Australia|copyrighted]] as it was generated by a [[software interface]] rather than a human author.<ref>{{Cite web|url=http://www.lavan.com.au/advice/intellectual_property_technology/copyright_in_source_code_and_digital_products|title=Copyright in source code and digital products|last=Lavan|website=Lavan|language=en|access-date=2019-05-10}}</ref> The same question may apply to neuromorphic systems: if a neuromorphic system successfully mimics a human brain and produces a piece of original work, who, if anyone, should be able to claim ownership of the work?<ref>{{cite journal |last1=Eshraghian|first1=Jason K. |title=Human Ownership of Artificial Creativity |journal=Nature Machine Intelligence |date=9 March 2020 |volume=2 |pages=157–160 |doi=10.1038/s42256-020-0161-x}}</ref><br />
<br />
There is significant legal debate around property rights and artificial intelligence. In Acohs Pty Ltd v. Ucorp Pty Ltd, Justice Christopher Jessup of the Federal Court of Australia found that the source code for Material Safety Data Sheets could not be copyrighted as it was generated by a software interface rather than a human author. The same question may apply to neuromorphic systems: if a neuromorphic system successfully mimics a human brain and produces a piece of original work, who, if anyone, should be able to claim ownership of the work?<br />
<br />
= = = 所有权和财产权 = = = 围绕财产权和人工智能有着重大的法律争论。在 Acohs Pty Ltd 诉 Ucorp Pty Ltd 一案中，澳大利亚联邦法院的克里斯托弗 · 杰瑟普法官发现，材料安全数据表的源代码不能受版权保护，因为它是由软件界面而不是人工作者生成的。同样的问题可能也适用于神经形态系统: 如果一个神经形态系统成功地模仿了人类的大脑并产生了一部原创作品，那么谁，如果有人，应该声称拥有这部作品的所有权？<br />
<br />
==Neuromemristive systems==<br />
Neuromemristive systems are a subclass of neuromorphic computing systems that focus on the use of [[memristors]] to implement [[neuroplasticity]]. While neuromorphic engineering focuses on mimicking biological behavior, neuromemristive systems focus on abstraction.<ref>{{Cite web|url=https://digitalops.sandia.gov/Mediasite/Play/a10cf6ceb55d47608bb8326dd00e46611d|title=002.08 N.I.C.E. Workshop 2014: Towards Intelligent Computing with Neuromemristive Circuits and Systems - Feb. 2014|website=digitalops.sandia.gov|access-date=2019-08-26}}</ref> For example, a neuromemristive system may replace the details of a [[Cerebral cortex|cortical]] microcircuit's behavior with an abstract neural network model.<ref>C. Merkel and D. Kudithipudi, "Neuromemristive extreme learning machines for pattern classification," ISVLSI, 2014.</ref><br />
<br />
Neuromemristive systems are a subclass of neuromorphic computing systems that focus on the use of memristors to implement neuroplasticity. While neuromorphic engineering focuses on mimicking biological behavior, neuromemristive systems focus on abstraction. For example, a neuromemristive system may replace the details of a cortical microcircuit's behavior with an abstract neural network model.C. Merkel and D. Kudithipudi, "Neuromemristive extreme learning machines for pattern classification," ISVLSI, 2014.<br />
<br />
神经记忆电阻系统是神经形态计算系统的一个亚类，主要研究利用记忆电阻器实现神经可塑性。神经形态工程的重点是模拟生物行为，而神经记忆电阻系统的重点是提取。例如，一个神经记忆系统可以用一个抽象的神经网络模型取代皮层微电路的行为细节。默克尔和 d. Kudithipudi，“用于模式分类的神经记忆极端学习机器”，ISVLSI，2014。<br />
<br />
There exist several neuron inspired threshold logic functions<ref name="Maan 1–13" /> implemented with memristors that have applications in high level [[pattern recognition]] applications. Some of the applications reported recently include [[speech recognition]],<ref>{{Cite journal|title = Memristor pattern recogniser: isolated speech word recognition|journal = Electronics Letters|pages = 1370–1372|volume = 51|issue = 17|doi = 10.1049/el.2015.1428|first1 = A.K.|last1 = Maan|first2 = A.P.|last2 = James|first3 = S.|last3 = Dimitrijev|year = 2015|bibcode = 2015ElL....51.1370M|hdl = 10072/140989|s2cid = 61454815|url = https://semanticscholar.org/paper/48d3ab11ec6e213b62f11eedcfb7b7febb058674|hdl-access = free}}</ref> [[face recognition]]<ref>{{Cite journal|title = Memristive Threshold Logic Face Recognition|journal = Procedia Computer Science|date = 2014-01-01|pages = 98–103|volume = 41|series = 5th Annual International Conference on Biologically Inspired Cognitive Architectures, 2014 BICA|doi = 10.1016/j.procs.2014.11.090|first1 = Akshay Kumar|last1 = Maan|first2 = Dinesh S.|last2 = Kumar|first3 = Alex Pappachen|last3 = James|doi-access = free}}</ref> and [[object recognition]].<ref>{{Cite journal|title = Memristive Threshold Logic Circuit Design of Fast Moving Object Detection|journal = IEEE Transactions on Very Large Scale Integration (VLSI) Systems|date = 2015-10-01|issn = 1063-8210|pages = 2337–2341|volume = 23|issue = 10|doi = 10.1109/TVLSI.2014.2359801|first1 = A.K.|last1 = Maan|first2 = D.S.|last2 = Kumar|first3 = S.|last3 = Sugathan|first4 = A.P.|last4 = James|arxiv = 1410.1267|s2cid = 9647290}}</ref> They also find applications in replacing conventional digital logic gates.<ref>{{Cite journal|title = Resistive Threshold Logic|journal = IEEE Transactions on Very Large Scale Integration (VLSI) Systems|date = 2014-01-01|issn = 1063-8210|pages = 190–195|volume = 22|issue = 1|doi = 10.1109/TVLSI.2012.2232946|first1 = A.P.|last1 = James|first2 = L.R.V.J.|last2 = Francis|first3 = D.S.|last3 = Kumar|arxiv = 1308.0090|s2cid = 7357110}}</ref><ref>{{Cite journal|title = Threshold Logic Computing: Memristive-CMOS Circuits for Fast Fourier Transform and Vedic Multiplication|journal = IEEE Transactions on Very Large Scale Integration (VLSI) Systems|date = 2015-11-01|issn = 1063-8210|pages = 2690–2694|volume = 23|issue = 11|doi = 10.1109/TVLSI.2014.2371857|first1 = A.P.|last1 = James|first2 = D.S.|last2 = Kumar|first3 = A.|last3 = Ajayan|arxiv = 1411.5255|s2cid = 6076956}}</ref><br />
<br />
There exist several neuron inspired threshold logic functions implemented with memristors that have applications in high level pattern recognition applications. Some of the applications reported recently include speech recognition, face recognition and object recognition. They also find applications in replacing conventional digital logic gates.<br />
<br />
利用记忆电阻器实现的神经元启发阈值逻辑函数在高级模式识别中有着广泛的应用。最近报道的一些应用包括语音识别、人脸识别和物体识别。它们还可以用来取代传统的数字逻辑门。<br />
<br />
For ideal passive memristive circuits there is an exact equation (Caravelli-Traversa-[[Di Ventra]] equation) for the internal memory of the circuit:<ref>{{cite journal |last=Caravelli |display-authors=etal|arxiv=1608.08651 |title=The complex dynamics of memristive circuits: analytical results and universal slow relaxation |year=2017 |doi=10.1103/PhysRevE.95.022140 |pmid= 28297937 |volume=95 |issue= 2 |pages= 022140 |journal=Physical Review E|bibcode=2017PhRvE..95b2140C |s2cid=6758362}}</ref><br />
<br />
For ideal passive memristive circuits there is an exact equation (Caravelli-Traversa-Di Ventra equation) for the internal memory of the circuit:<br />
<br />
对于理想的无源记忆电路，电路的内部记忆有一个精确的方程(Caravelli-Traversa-Di Ventra 方程) :<br />
<br />
:<math> \frac{d}{dt} \vec{W} = \alpha \vec{W}-\frac{1}{\beta} (I+\xi \Omega W)^{-1} \Omega \vec S </math><br />
<br />
:\frac{d}{dt} \vec{W} = \alpha \vec{W}-\frac{1}{\beta} (I+\xi \Omega W)^{-1} \Omega \vec S<br />
<br />
:\frac{d}{dt} \vec{W} = \alpha \vec{W}-\frac{1}{\beta} (I+\xi \Omega W)^{-1} \Omega \vec S<br />
<br />
as a function of the properties of the physical memristive network and the external sources. In the equation above, <math>\alpha</math> is the "forgetting" time scale constant, <math> \xi=r-1</math> and <math>r=\frac{R_\text{off}}{R_\text{on}}</math> is the ratio of ''off'' and ''on'' values of the limit resistances of the memristors, <math> \vec S </math> is the vector of the sources of the circuit and <math>\Omega</math> is a projector on the fundamental loops of the circuit. The constant <math>\beta</math> has the dimension of a voltage and is associated to the properties of the [[memristor]]; its physical origin is the charge mobility in the conductor. The diagonal matrix and vector <math>W=\operatorname{diag}(\vec W)</math> and <math>\vec W</math> respectively, are instead the internal value of the memristors, with values between 0 and 1. This equation thus requires adding extra constraints on the memory values in order to be reliable.<br />
<br />
<nowiki>as a function of the properties of the physical memristive network and the external sources. In the equation above, \alpha is the "forgetting" time scale constant, \xi=r-1 and r=\frac{R_\text{off}}{R_\text{on}} is the ratio of off and on values of the limit resistances of the memristors, \vec S is the vector of the sources of the circuit and \Omega is a projector on the fundamental loops of the circuit. The constant \beta has the dimension of a voltage and is associated to the properties of the memristor; its physical origin is the charge mobility in the conductor. The diagonal matrix and vector W=\operatorname{diag}(\vec W) and \vec W respectively, are instead the internal value of the memristors, with values between 0 and 1. This equation thus requires adding extra constraints on the memory values in order to be reliable.</nowiki><br />
<br />
<nowiki>作为物理记忆网络和外部源的性质的函数。在上述方程中，α 是“遗忘”时间尺度常数，xi = r-1，r = frac { r _ text { off }{ on }{ r _ text { on }}是记忆电阻器极限电阻的开关和开关值之比，vec s 是电路源的矢量，Omega 是电路基本环路的投影仪。常数 β 具有电压的尺寸，与记忆电阻器的特性有关; 它的物理起源是导体中的电荷迁移率。对角矩阵和向量 w = 操作者名{ diag }(vec w)和 vec w 分别是记忆电阻器的内值，值在0到1之间。因此，这个等式需要在内存值上添加额外的约束，以保证可靠性。</nowiki><br />
<br />
==See also==<br />
{{Columns-list|colwidth=18em|<br />
* [[AI accelerator (computer hardware)|AI accelerator]]<br />
* [[Artificial brain]]<br />
* [[Biomorphic]]<br />
* [[Cognitive computer]]<br />
* [[Computation and Neural Systems]]<br />
* [[Differentiable programming]]<br />
* [[Event camera]]<br />
* [[Neurorobotics]]<br />
* [[Optical flow sensor]]<br />
* [[Physical neural network]]<br />
* [[SpiNNaker]]<br />
* [[SyNAPSE]]<br />
* [[Retinomorphic sensor]]<br />
* [[Vision chip]]<br />
* [[Vision processing unit]]<br />
* [[Zeroth (software)]]<br />
* [[Hardware for artificial intelligence]]<br />
}}<br />
<br />
{{Portal bar|Electronics}}<br />
<br />
== References==<br />
{{Reflist|40em}}<br />
<br />
==External links==<br />
<!--======================== {{No more links}} ============================<br />
| PLEASE BE CAUTIOUS IN ADDING MORE LINKS TO THIS ARTICLE. Wikipedia |<br />
| is not a collection of links nor should it be used for advertising. |<br />
| |<br />
| Excessive or inappropriate links WILL BE DELETED. |<br />
| See [[Wikipedia:External links]] & [[Wikipedia:Spam]] for details. |<br />
| |<br />
| If there are already plentiful links, please propose additions or |<br />
| replacements on this article's discussion page, or submit your link |<br />
| to the relevant category at the Open Directory Project (dmoz.org) |<br />
| and link back to that category using the {{dmoz}} template. |<br />
======================={{No more links}}=============================--><br />
{{Commons category|Neuromorphic Engineering}}<br />
*[https://web.archive.org/web/20150727034331/http://ine-web.org/workshops/workshops-overview Telluride Neuromorphic Engineering Workshop]<br />
*[https://archive.today/20130115190057/http://capocaccia.ethz.ch/ CapoCaccia Cognitive Neuromorphic Engineering Workshop]<br />
*[https://web.archive.org/web/20190716132350/http://www.ine-web.org/ Institute of Neuromorphic Engineering]<br />
*[http://www.ine-news.org/ INE news site].<br />
*[http://www.frontiersin.org/neuromorphic_engineering Frontiers in Neuromorphic Engineering Journal]<br />
*[http://www.cns.caltech.edu/ Computation and Neural Systems] department at the [[California Institute of Technology]].<br />
*[http://www.humanbrainproject.eu/ Human Brain Project official site]<br />
*[https://www.the-scientist.com/features/building-a-silicon-brain-65738 Building a Silicon Brain:] Computer chips based on biological neurons may help simulate larger and more-complex brain models. May 1, 2019. SANDEEP RAVINDRAN<br />
<br />
<br />
*Telluride Neuromorphic Engineering Workshop<br />
*CapoCaccia Cognitive Neuromorphic Engineering Workshop<br />
*Institute of Neuromorphic Engineering<br />
*INE news site.<br />
*Frontiers in Neuromorphic Engineering Journal<br />
*Computation and Neural Systems department at the California Institute of Technology.<br />
*Human Brain Project official site<br />
*Building a Silicon Brain: Computer chips based on biological neurons may help simulate larger and more-complex brain models. May 1, 2019. SANDEEP RAVINDRAN<br />
<br />
=<nowiki>= = 外部链接 =</nowiki>= <br />
*碲化物神经形态工程工作室<br />
*CapoCaccia 认知神经形态工程工作室<br />
*神经形态工程研究所<br />
* INE 新闻站点。<br />
*《神经形态工程学前沿》<br />
* 加州理工学院计算与神经系统系。<br />
*人脑项目官方网站<br />
*打造硅脑: 基于生物神经元的计算机芯片可能有助于模拟更大、更复杂的大脑模型。2019年5月1日。SANDEEP RAVINDRAN<br />
<br />
{{Differentiable computing}}<br />
{{Authority control}}<br />
<br />
[[Category:Electrical engineering]]<br />
[[Category:Neuroscience]]<br />
[[分类:AI accelerators|*]]<br />
[[Category:Artificial intelligence]]<br />
[[Category:Robotics]]<br />
<br />
Category:Electrical engineering<br />
Category:Neuroscience<br />
*<br />
Category:Artificial intelligence<br />
Category:Robotics<br />
<br />
类别: 电气工程类别: 神经科学 * 类别: 人工智能类别: 机器人<br />
<br />
<noinclude><br />
<br />
<small>This page was moved from [[wikipedia:en:Neuromorphic engineering]]. Its edit history can be viewed at [[神经形态计算/edithistory]]</small></noinclude><br />
<br />
[[Category:待整理页面]]<br />
<br />
<references /></div>Shenky20https://wiki.swarma.org/index.php?title=%E7%A5%9E%E7%BB%8F%E5%BD%A2%E6%80%81%E8%AE%A1%E7%AE%97&diff=30171神经形态计算2022-04-13T07:08:19Z<p>Shenky20：</p>
<hr />
<div><br />
此词条由神经动力学模型读书会词条梳理志愿者Shenky20翻译审校，未经专家审校，带来阅读不便，请见谅。<br />
<br />
{{Use American English|date = January 2019}}<br />
{{Use mdy dates|date = January 2019}}<br />
<br />
'''Neuromorphic engineering''', also known as '''neuromorphic computing''',<ref>{{Cite journal | doi = 10.1145/2601069| title = Neuromorphic computing gets ready for the (really) big time| journal = [[Communications of the ACM]]| volume = 57| issue = 6| year = 2014| last1 = Monroe | first1 = D. | pages = 13–15| s2cid = 20051102}}</ref><ref>{{Cite journal | doi = 10.1088/0957-4484/21/17/175202| pmid = 20368686| title = Nanotube devices based crossbar architecture: Toward neuromorphic computing| journal = Nanotechnology| volume = 21| issue = 17| pages = 175202| year = 2010| last1 = Zhao | first1 = W. S. | last2 = Agnus | first2 = G. | last3 = Derycke | first3 = V. | last4 = Filoramo | first4 = A. | last5 = Bourgoin | first5 = J. -P. | last6 = Gamrat | first6 = C. | bibcode = 2010Nanot..21q5202Z| url = https://zenodo.org/record/3428659}}</ref><ref name="humanbrainproject">{{YouTube|id=6RoiZ90mGfw|title=The Human Brain Project SP 9: Neuromorphic Computing Platform}}</ref> is the use of [[very-large-scale integration]] (VLSI) systems containing electronic [[analog circuit]]s to mimic neuro-biological architectures present in the nervous system. A neuromorphic computer/chip is any device that uses physical artificial neurons (made from silicon) to do computations.<ref>{{cite journal|last1=Mead|first1=Carver|title=Neuromorphic electronic systems|journal=Proceedings of the IEEE|date=1990|volume=78|issue=10|pages=1629–1636|doi=10.1109/5.58356|url=https://authors.library.caltech.edu/53090/1/00058356.pdf}}</ref><ref name=":2" /> In recent times, the term ''neuromorphic'' has been used to describe analog, digital, mixed-mode analog/digital VLSI, and software systems that implement models of [[neural system]]s (for [[perception]], [[motor control]], or [[multisensory integration]]). The implementation of neuromorphic computing on the hardware level can be realized by oxide-based [[memristor]]s,<ref name="Maan 1–13">{{Cite journal|last1=Maan|first1=A. K.|last2=Jayadevi|first2=D. A.|last3=James|first3=A. P.|date=2016-01-01|title=A Survey of Memristive Threshold Logic Circuits|journal=IEEE Transactions on Neural Networks and Learning Systems|volume=PP|issue=99|pages=1734–1746|doi=10.1109/TNNLS.2016.2547842|pmid=27164608|issn=2162-237X|arxiv=1604.07121|bibcode=2016arXiv160407121M|s2cid=1798273}}</ref> spintronic memories, threshold switches, and [[transistor]]s.<ref>{{Cite journal|title = Mott Memory and Neuromorphic Devices|journal = Proceedings of the IEEE|date = 2015-08-01|issn = 0018-9219|pages = 1289–1310|volume = 103|issue = 8|doi = 10.1109/JPROC.2015.2431914|first1 = You|last1 = Zhou|first2 = S.|last2 = Ramanathan|s2cid = 11347598|url = https://zenodo.org/record/895565}}</ref><ref name=":2">{{Cite document|title=Neuromorphic Circuits With Neural Modulation Enhancing the Information Content of Neural Signaling {{!}} International Conference on Neuromorphic Systems 2020|language=EN|doi=10.1145/3407197.3407204|s2cid=220794387}}</ref> Training software-based neuromorphic systems of spiking neural networks can be achieved using error backpropagation, e.g., using [[Python (programming language)|Python]] based frameworks such as snnTorch,<ref>{{cite journal |last1=Eshraghian|first1=Jason K.|last2=Ward|first2=Max|last3=Neftci |first3=Emre|last4=Wang|first4=Xinxin|last5=Lenz|first5=Gregor|last6=Dwivedi|first6=Girish|last7=Bennamoun|first7=Mohammed|last8=Jeong|first8=Doo Seok|last9=Lu|first9=Wei D.|title=Training Spiking Neural Networks Using Lessons from Deep Learning |date=1 October 2021 |arxiv=2109.12894 }}</ref> or using canonical learning rules from the biological learning literature, e.g., using BindsNet.<ref>{{Cite web | url=https://github.com/Hananel-Hazan/bindsnet | title=Hananel-Hazan/bindsnet: Simulation of spiking neural networks (SNNs) using PyTorch.| date=31 March 2020}}</ref><br />
<br />
Neuromorphic engineering, also known as neuromorphic computing, is the use of very-large-scale integration (VLSI) systems containing electronic analog circuits to mimic neuro-biological architectures present in the nervous system. A neuromorphic computer/chip is any device that uses physical artificial neurons (made from silicon) to do computations. In recent times, the term neuromorphic has been used to describe analog, digital, mixed-mode analog/digital VLSI, and software systems that implement models of neural systems (for perception, motor control, or multisensory integration). The implementation of neuromorphic computing on the hardware level can be realized by oxide-based memristors, spintronic memories, threshold switches, and transistors. Training software-based neuromorphic systems of spiking neural networks can be achieved using error backpropagation, e.g., using Python based frameworks such as snnTorch, or using canonical learning rules from the biological learning literature, e.g., using BindsNet.<br />
<br />
'''<font color="#ff8000">神经形态工程Neuromorphic engineering</font>'''(也称为'''<font color="#ff8000">神经形态计算Neuromorphic computing</font>''')是指使用包含电子'''<font color="#ff8000">模拟电路Analog circuit</font>'''的'''<font color="#ff8000">超大规模集成电路Very-large-scale integration</font>'''系统来模拟神经系统中生理结构的研究方法。神经形态计算机或神经形态芯片包括任何使用由硅制成的人造神经元进行计算的设备。近年来，神经形态学（neuromorphic）这个术语被用来描述能够实现'''<font color="#ff8000">神经系统Neural system</font>'''模型功能(如'''<font color="#ff8000">感知Perception</font>'''、'''<font color="#ff8000">运动控制Motor control</font>'''，'''<font color="#ff8000">多感官整合Multisensory integration</font>'''等)的模拟、数字、模拟/数字混合模式超大规模集成电路和软件系统。神经形态计算的硬件实现可以通过基于氧化物的'''<font color="#ff8000">记忆电阻器Memristor</font>'''、自旋电子存储器、阈值开关和'''<font color="#ff8000">晶体管Transistor</font>'''来实现。对基于软件的脉冲神经网络系统的训练可以通过误差反向传播机制来实现，例如，使用snnTorch等基于Python的框架，或使用BindsNet等典型的受生物启发的学习模式。<br />
<br />
A key aspect of neuromorphic engineering is understanding how the morphology of individual neurons, circuits, applications, and overall architectures creates desirable computations, affects how information is represented, influences robustness to damage, incorporates learning and development, adapts to local change (plasticity), and facilitates evolutionary change.<br />
<br />
A key aspect of neuromorphic engineering is understanding how the morphology of individual neurons, circuits, applications, and overall architectures creates desirable computations, affects how information is represented, influences robustness to damage, incorporates learning and development, adapts to local change (plasticity), and facilitates evolutionary change.<br />
<br />
神经形态工程领域的一个关键问题，就是理解'''<font color="#32CD32">单个神经元形态、神经回路、应用和整体结构</font>'''如何产生理想的计算，如何影响信息的表示和对破坏的鲁棒性，如何整合学习和发展，如何适应局部变化(可塑性) 并促进逐渐发展的变化。<br />
<br />
Neuromorphic engineering is an interdisciplinary subject that takes inspiration from [[biology]], [[physics]], [[mathematics]], [[computer science]], and [[electronic engineering]]<ref name=":2" /> to design artificial neural systems, such as [[Machine vision|vision systems]], head-eye systems, auditory processors, and autonomous robots, whose physical architecture and design principles are based on those of biological nervous systems.<ref>{{Cite journal | doi = 10.1155/2012/705483| title = Qualitative Functional Decomposition Analysis of Evolved Neuromorphic Flight Controllers| journal = Applied Computational Intelligence and Soft Computing| volume = 2012| pages = 1–21| year = 2012| last1 = Boddhu | first1 = S. K. | last2 = Gallagher | first2 = J. C. | doi-access = free}}</ref> It was developed by [[Carver Mead]]<ref>{{cite web|title=carver mead website|last1=Mead|first1=Carver|url=http://www.carvermead.caltech.edu/index.html|website=carvermead}}</ref> in the late 1980s.<br />
<br />
Neuromorphic engineering is an interdisciplinary subject that takes inspiration from biology, physics, mathematics, computer science, and electronic engineering to design artificial neural systems, such as vision systems, head-eye systems, auditory processors, and autonomous robots, whose physical architecture and design principles are based on those of biological nervous systems. It was developed by Carver Mead in the late 1980s.<br />
<br />
神经形态工程是以生物学、物理学、数学、计算机科学和电子工程为基础，设计人工神经系统，如视觉系统、头眼系统、听觉处理器和自主机器人的一门交叉学科。它是由卡弗 · 米德在20世纪80年代后期开发的。<br />
<br />
== Neurological inspiration ==<br />
Neuromorphic engineering is set apart by the inspiration it takes from what we know about the structure and operations of the [[brain]]. Neuromorphic engineering translates what we know about the brain's function into computer systems. Work has mostly focused on replicating the analog nature of [[biological computation]] and the role of [[neuron]]s in [[cognition]].<br />
<br />
Neuromorphic engineering is set apart by the inspiration it takes from what we know about the structure and operations of the brain. Neuromorphic engineering translates what we know about the brain's function into computer systems. Work has mostly focused on replicating the analog nature of biological computation and the role of neurons in cognition.<br />
<br />
神经学灵感工程学是从我们对大脑结构和运作的了解中获得灵感而形成的。神经形态工程学将我们对大脑功能的了解转化为计算机系统。工作主要集中在复制生物计算的模拟特性和神经元在认知中的作用。<br />
<br />
The biological processes of neurons and their [[synapse]]s are dauntingly complex, and thus very difficult to artificially simulate. A key feature of biological brains is that all of the processing in neurons use analog chemical signals. This makes it hard to replicate brains in computers because the current generation of computers is completely digital. However, the characteristics of these parts can be abstracted into mathematical functions that closely capture the essence of the neuron's operations.<br />
<br />
The biological processes of neurons and their synapses are dauntingly complex, and thus very difficult to artificially simulate. A key feature of biological brains is that all of the processing in neurons use analog chemical signals. This makes it hard to replicate brains in computers because the current generation of computers is completely digital. However, the characteristics of these parts can be abstracted into mathematical functions that closely capture the essence of the neuron's operations.<br />
<br />
神经元及其突触的生物学过程极其复杂，难以人工模拟。生物大脑的一个关键特征是，神经元的所有处理过程都使用模拟化学信号。这使得很难在计算机中复制大脑，因为目前的计算机是完全数字化的。然而，这些部分的特征可以抽象为数学函数，这些函数紧密地抓住了神经元运行的本质。<br />
<br />
The goal of neuromorphic computing is not to perfectly mimic the brain and all of its functions, but instead to extract what is known of its structure and operations to be used in a practical computing system. No neuromorphic system will claim nor attempt to reproduce every element of neurons and synapses, but all adhere to the idea that computation is highly [[distributed processing|distributed]] throughout a series of small computing elements analogous to a neuron. While this sentiment is standard, researchers chase this goal with different methods.<ref>{{Cite journal | doi = 10.1088/1741-2560/13/5/051001| title = Large-scale neuromorphic computing systems| journal = Journal of Neural Engineering| volume = 13| pages = 1–15| year = 2016| last1 = Furber | first1 = Steve| issue = 5| pmid = 27529195| bibcode = 2016JNEng..13e1001F| doi-access = free}}</ref><br />
<br />
The goal of neuromorphic computing is not to perfectly mimic the brain and all of its functions, but instead to extract what is known of its structure and operations to be used in a practical computing system. No neuromorphic system will claim nor attempt to reproduce every element of neurons and synapses, but all adhere to the idea that computation is highly distributed throughout a series of small computing elements analogous to a neuron. While this sentiment is standard, researchers chase this goal with different methods.<br />
<br />
神经形态计算的目标不是完美地模拟大脑及其所有功能，而是提取已知的大脑结构和操作，用于实际的计算系统。没有哪个神经形态学系统会声称或试图复制神经元和突触的每一个元素，但所有人都坚持这样的观点，即计算是高度分布在一系列类似于神经元的小型计算元素中的。虽然这种情绪是标准的，但研究人员用不同的方法追求这一目标。<br />
<br />
== Examples==<br />
As early as 2006, researchers at [[Georgia Tech]] published a field programmable neural array.<ref>{{Cite book|title = A field programmable neural array|last1 = Farquhar|first1 = Ethan|date = May 2006|journal = IEEE International Symposium on Circuits and Systems|pages = 4114–4117|last2 = Hasler|first2 = Paul.|doi = 10.1109/ISCAS.2006.1693534|isbn = 978-0-7803-9389-9|s2cid = 206966013}}</ref> This chip was the first in a line of increasingly complex arrays of floating gate transistors that allowed programmability of charge on the gates of [[MOSFET]]s to model the channel-ion characteristics of neurons in the brain and was one of the first cases of a silicon programmable array of neurons.<br />
<br />
As early as 2006, researchers at Georgia Tech published a field programmable neural array. This chip was the first in a line of increasingly complex arrays of floating gate transistors that allowed programmability of charge on the gates of MOSFETs to model the channel-ion characteristics of neurons in the brain and was one of the first cases of a silicon programmable array of neurons.<br />
<br />
早在2006年，佐治亚理工学院的研究人员就发表了现场可编程神经阵列。这种芯片是一系列越来越复杂的浮栅晶体管阵列中的第一个，这些晶体管可以在 mosfet 的栅极上编程来模拟大脑中神经元的通道离子特性，这也是硅可编程神经元阵列的首批案例之一。<br />
<br />
In November 2011, a group of [[MIT]] researchers created a computer chip that mimics the analog, ion-based communication in a synapse between two neurons using 400 transistors and standard [[CMOS]] manufacturing techniques.<ref>{{cite web|title=MIT creates "brain chip"|url=http://www.extremetech.com/extreme/105067-mit-creates-brain-chip|access-date=4 December 2012}}</ref><ref name="Neuromorphic silicon paper">{{cite journal|title=Neuromorphic silicon neurons and large-scale neural networks: challenges and opportunities|doi=10.3389/fnins.2011.00108|pmid=21991244|pmc=3181466|volume=5|pages=108|journal=Frontiers in Neuroscience|year=2011|last1=Poon|first1=Chi-Sang|last2=Zhou|first2=Kuan|doi-access=free}}</ref><br />
<br />
In November 2011, a group of MIT researchers created a computer chip that mimics the analog, ion-based communication in a synapse between two neurons using 400 transistors and standard CMOS manufacturing techniques.<br />
<br />
2011年11月，麻省理工学院的一组研究人员发明了一种计算机芯片，该芯片使用400个晶体管和标准的 CMOS 制造技术，在两个神经元之间的突触中模拟模拟基于离子的通讯。<br />
<br />
In June 2012, [[spintronic]] researchers at [[Purdue University]] presented a paper on the design of a neuromorphic chip using [[Spin valve|lateral spin valve]]s and [[memristor]]s. They argue that the architecture works similarly to neurons and can therefore be used to test methods of reproducing the brain's processing. In addition, these chips are significantly more energy-efficient than conventional ones.<ref name="Spin Devices Prop">{{Cite arXiv|title=Proposal For Neuromorphic Hardware Using Spin Devices|eprint=1206.3227|last1=Sharad|first1=Mrigank|last2=Augustine|first2=Charles|last3=Panagopoulos|first3=Georgios|last4=Roy|first4=Kaushik|class=cond-mat.dis-nn|year=2012}}</ref><br />
<br />
In June 2012, spintronic researchers at Purdue University presented a paper on the design of a neuromorphic chip using lateral spin valves and memristors. They argue that the architecture works similarly to neurons and can therefore be used to test methods of reproducing the brain's processing. In addition, these chips are significantly more energy-efficient than conventional ones.<br />
<br />
2012年6月，普渡大学的自旋电子学研究人员发表了一篇关于利用侧向自旋阀和记忆电阻器设计神经形态芯片的论文。他们认为，这种结构的工作原理与神经元相似，因此可以用来测试复制大脑处理过程的方法。此外，这些芯片明显比传统芯片更节能。<br />
<br />
Research at [[HP Labs]] on Mott memristors has shown that while they can be non-[[Volatile memory|volatile]], the volatile behavior exhibited at temperatures significantly below the [[phase transition]] temperature can be exploited to fabricate a [[neuristor]],<ref name=":0" /> a biologically-inspired device that mimics behavior found in neurons.<ref name=":0">{{Cite journal | doi = 10.1038/nmat3510| pmid = 23241533| title = A scalable neuristor built with Mott memristors| journal = Nature Materials| volume = 12| issue = 2| pages = 114–7| year = 2012| last1 = Pickett | first1 = M. D. | last2 = Medeiros-Ribeiro | first2 = G. | last3 = Williams | first3 = R. S. | bibcode = 2013NatMa..12..114P| s2cid = 16271627| url = https://semanticscholar.org/paper/b6ba6f496ace2b947f111059663e76bb60e9efeb}}</ref> In September 2013, they presented models and simulations that show how the spiking behavior of these neuristors can be used to form the components required for a [[Turing machine]].<ref>{{cite journal|doi=10.1088/0957-4484/24/38/384002|title=Phase transitions enable computational universality in neuristor-based cellular automata|author1=Matthew D Pickett|author2=R Stanley Williams|name-list-style=amp|date=September 2013|publisher=IOP Publishing Ltd|journal=Nanotechnology|volume=24|issue=38|pmid=23999059|bibcode=2013Nanot..24L4002P|at=384002}}</ref><br />
<br />
Research at HP Labs on Mott memristors has shown that while they can be non-volatile, the volatile behavior exhibited at temperatures significantly below the phase transition temperature can be exploited to fabricate a neuristor, a biologically-inspired device that mimics behavior found in neurons. In September 2013, they presented models and simulations that show how the spiking behavior of these neuristors can be used to form the components required for a Turing machine.<br />
<br />
惠普实验室在莫特记忆电阻器上的研究表明，尽管它们可以是非挥发性的，但是在相变温度以下显示的挥发性行为可以被用来制造神经元电阻器，这是一种模仿神经元行为的生物设备。2013年9月，他们展示了模型和仿真，展示了这些神经元的尖峰行为是如何被用来形成图灵机所需的元件的。<br />
<br />
[[Neurogrid]], built by ''Brains in Silicon'' at [[Stanford University]],<ref>{{cite journal|last1=Boahen|first1=Kwabena|title=Neurogrid: A Mixed-Analog-Digital Multichip System for Large-Scale Neural Simulations|journal=Proceedings of the IEEE|date=24 April 2014|volume=102|issue=5|pages=699–716|doi=10.1109/JPROC.2014.2313565|s2cid=17176371}}</ref> is an example of hardware designed using neuromorphic engineering principles. The circuit board is composed of 16 custom-designed chips, referred to as NeuroCores. Each NeuroCore's analog circuitry is designed to emulate neural elements for 65536 neurons, maximizing energy efficiency. The emulated neurons are connected using digital circuitry designed to maximize spiking throughput.<ref>{{cite journal|doi=10.1038/503022a|pmid = 24201264|title = Neuroelectronics: Smart connections|journal = Nature|volume = 503|issue = 7474|pages = 22–4|year = 2013|last1 = Waldrop|first1 = M. Mitchell|bibcode = 2013Natur.503...22W|doi-access = free}}</ref><ref>{{cite journal|doi=10.1109/JPROC.2014.2313565|title = Neurogrid: A Mixed-Analog-Digital Multichip System for Large-Scale Neural Simulations|journal = Proceedings of the IEEE|volume = 102|issue = 5|pages = 699–716|year = 2014|last1 = Benjamin|first1 = Ben Varkey|last2 = Peiran Gao|last3 = McQuinn|first3 = Emmett|last4 = Choudhary|first4 = Swadesh|last5 = Chandrasekaran|first5 = Anand R.|last6 = Bussat|first6 = Jean-Marie|last7 = Alvarez-Icaza|first7 = Rodrigo|last8 = Arthur|first8 = John V.|last9 = Merolla|first9 = Paul A.|last10 = Boahen|first10 = Kwabena|s2cid = 17176371}}</ref><br />
<br />
Neurogrid, built by Brains in Silicon at Stanford University, is an example of hardware designed using neuromorphic engineering principles. The circuit board is composed of 16 custom-designed chips, referred to as NeuroCores. Each NeuroCore's analog circuitry is designed to emulate neural elements for 65536 neurons, maximizing energy efficiency. The emulated neurons are connected using digital circuitry designed to maximize spiking throughput.<br />
<br />
神经网格是由斯坦福大学硅谷大脑公司建立的，是一个使用神经形态工程原理设计硬件的例子。该电路板由16个定制设计的芯片组成，称为 NeuroCores。每个 NeuroCore 的模拟电路被设计为模拟65536个神经元的神经元元件，最大限度地提高能量效率。模拟的神经元通过设计的数字电路连接，以最大化脉冲吞吐量。<br />
<br />
A research project with implications for neuromorphic engineering is the [[Human Brain Project]] that is attempting to simulate a complete human brain in a supercomputer using biological data. It is made up of a group of researchers in neuroscience, medicine, and computing.<ref>{{cite web|title=Involved Organizations|url=http://www.humanbrainproject.eu/partners.html|access-date=22 February 2013|url-status=dead|archive-url=https://web.archive.org/web/20130302142627/http://www.humanbrainproject.eu/partners.html|archive-date=2 March 2013}}</ref> [[Henry Markram]], the project's co-director, has stated that the project proposes to establish a foundation to explore and understand the brain and its diseases, and to use that knowledge to build new computing technologies. The three primary goals of the project are to better understand how the pieces of the brain fit and work together, to understand how to objectively diagnose and treat brain diseases, and to use the understanding of the human brain to develop neuromorphic computers. That the simulation of a complete human brain will require a supercomputer a thousand times more powerful than today's encourages the current focus on neuromorphic computers.<ref>{{cite web|title=Human Brain Project|url=http://www.humanbrainproject.eu|access-date=22 February 2013}}</ref> $1.3 billion has been allocated to the project by The [[European Commission]].<ref>{{cite web|title=The Human Brain Project and Recruiting More Cyberwarriors|url=http://www.marketplace.org/topics/tech/human-brain-project-and-recruiting-more-cyberwarriors|access-date=22 February 2013|date=January 29, 2013}}</ref><br />
<br />
A research project with implications for neuromorphic engineering is the Human Brain Project that is attempting to simulate a complete human brain in a supercomputer using biological data. It is made up of a group of researchers in neuroscience, medicine, and computing. Henry Markram, the project's co-director, has stated that the project proposes to establish a foundation to explore and understand the brain and its diseases, and to use that knowledge to build new computing technologies. The three primary goals of the project are to better understand how the pieces of the brain fit and work together, to understand how to objectively diagnose and treat brain diseases, and to use the understanding of the human brain to develop neuromorphic computers. That the simulation of a complete human brain will require a supercomputer a thousand times more powerful than today's encourages the current focus on neuromorphic computers. $1.3 billion has been allocated to the project by The European Commission.<br />
<br />
一个对神经形态工程有影响的研究项目是人脑项目，它试图用生物数据在超级计算机中模拟完整的人脑。它由一群神经科学、医学和计算机科学的研究人员组成。该项目的联合主管亨利•马克拉姆(Henry Markram)表示，该项目提议建立一个基础，以探索和了解大脑及其疾病，并利用这些知识构建新的计算机技术。这个项目的三个主要目标是: 更好地理解大脑的各个部分是如何相互匹配和协同工作的; 理解如何客观地诊断和治疗脑部疾病; 以及利用对人类大脑的理解来开发神经形态计算机。模拟一个完整的人类大脑需要一台比现在强大一千倍的超级计算机，这鼓励了当前对神经形态计算机的关注。欧洲委员会已经拨款13亿美元用于这个项目。<br />
<br />
Other research with implications for neuromorphic engineering involves the [[BRAIN Initiative]]<ref name="economist">[https://www.economist.com/news/science-and-technology/21582495-computers-will-help-people-understand-brains-better-and-understanding-brains Neuromorphic computing: The machine of a new soul], The Economist, 2013-08-03</ref> and the [[TrueNorth]] chip from [[IBM]].<ref>{{cite journal|last1=Modha|first1=Dharmendra|title=A million spiking-neuron integrated circuit with a scalable communication network and interface|journal=Science|date=Aug 2014|volume=345|issue=6197|pages=668–673|doi=10.1126/science.1254642|pmid=25104385|bibcode=2014Sci...345..668M|s2cid=12706847}}</ref> Neuromorphic devices have also been demonstrated using nanocrystals, nanowires, and conducting polymers.<ref>{{Cite web|url=http://jessamynfairfield.com/wp-content/uploads/2017/03/PWMar17Fairfield.pdf|title=Smarter Machines|last=Fairfield|first=Jessamyn|date=March 1, 2017}}</ref><br />
<br />
Other research with implications for neuromorphic engineering involves the BRAIN InitiativeNeuromorphic computing: The machine of a new soul, The Economist, 2013-08-03 and the TrueNorth chip from IBM. Neuromorphic devices have also been demonstrated using nanocrystals, nanowires, and conducting polymers.<br />
<br />
其他与神经形态工程有关的研究还包括 BRAIN initiativity/euromorphic computing: The machine of a new soul，The Economist，2013-08-03 and The TrueNorth chip from IBM。神经形态学设备也已经被证明使用纳米晶体、纳米线和导电聚合物。<br />
<br />
[[Intel]] unveiled its neuromorphic research chip, called “[[Intel Loihi|Loihi]]”, in October 2017. The chip uses an asynchronous [[spiking neural network]] (SNN) to implement adaptive self-modifying event-driven fine-grained parallel computations used to implement learning and inference with high efficiency.<ref>{{cite journal |last1=Davies |first1=Mike |title=Loihi: A Neuromorphic Manycore Processor with On-Chip Learning |journal=IEEE Micro |date=January 16, 2018 |volume=38 |issue=1 |pages=82–99 |display-authors=etal|doi=10.1109/MM.2018.112130359 |s2cid=3608458 }}</ref><ref name="Morris2017">{{cite web |last1=Morris |first1=John |title=Why Intel built a neuromorphic chip |url=https://www.zdnet.com/article/why-intel-built-a-neuromorphic-chip/ |website=ZDNet |access-date=17 August 2018 |language=en}}</ref><br />
<br />
Intel unveiled its neuromorphic research chip, called “Loihi”, in October 2017. The chip uses an asynchronous spiking neural network (SNN) to implement adaptive self-modifying event-driven fine-grained parallel computations used to implement learning and inference with high efficiency.<br />
<br />
2017年10月，英特尔公布了其神经形态研究芯片“ Loihi”。该芯片采用异步脉冲神经网络(SNN)实现自适应自修改事件驱动的细粒度并行计算，实现了高效的学习和推理。<br />
<br />
[[IMEC]], a Belgium-based nanoelectronics research center, demonstrated the world's first self-learning neuromorphic chip. The brain-inspired chip, based on OxRAM technology, has the capability of self-learning and has been demonstrated to have the ability to compose music.<ref>{{cite web |title=Imec demonstrates self-learning neuromorphic chip that composes music |url=https://www.imec-int.com/en/articles/imec-demonstrates-self-learning-neuromorphic-chip-that-composes-music |website=IMEC International |access-date=1 October 2019}}</ref> IMEC released the 30-second tune composed by the prototype. The chip was sequentially loaded with songs in the same time signature and style. The songs were old Belgian and French flute minuets, from which the chip learned the rules at play and then applied them.<ref>{{cite web|last1=Bourzac|first1=Katherine|title=A Neuromorphic Chip That Makes Music|url=https://spectrum.ieee.org/a-neuromorphic-chip-that-makes-music|url-status=live|access-date=1 October 2019|website=IEEE Spectrum|date=May 23, 2017}}</ref><br />
<br />
IMEC, a Belgium-based nanoelectronics research center, demonstrated the world's first self-learning neuromorphic chip. The brain-inspired chip, based on OxRAM technology, has the capability of self-learning and has been demonstrated to have the ability to compose music. IMEC released the 30-second tune composed by the prototype. The chip was sequentially loaded with songs in the same time signature and style. The songs were old Belgian and French flute minuets, from which the chip learned the rules at play and then applied them.<br />
<br />
比利时的一个纳米电子学研究中心 IMEC 展示了世界上第一个自学习神经形态芯片。这种基于 OxRAM 技术的大脑启发芯片具有自学习能力，并已被证明具有创作音乐的能力。IMEC 发布了由原型机谱写的30秒曲调。芯片按顺序加载同时签名和风格的歌曲。歌曲是古老的比利时和法国长笛小步舞曲，筹码从中学习游戏规则，然后应用它们。<br />
<br />
[[Blue Brain Project|The Blue Brain Project]], led by Henry Markram, aims to build biologically detailed digital reconstructions and simulations of the mouse brain. The Blue Brain Project has created in silico models of rodent brains, while attempting to replicate as many details about its biology as possible. The supercomputer-based simulations offer new perspectives on understanding the structure and functions of the brain.<br />
<br />
The Blue Brain Project, led by Henry Markram, aims to build biologically detailed digital reconstructions and simulations of the mouse brain. The Blue Brain Project has created in silico models of rodent brains, while attempting to replicate as many details about its biology as possible. The supercomputer-based simulations offer new perspectives on understanding the structure and functions of the brain.<br />
<br />
由亨利 · 马克拉姆领导的蓝脑计划旨在建立小鼠大脑的生物细节数字重建和模拟。蓝色大脑项目已经在啮齿动物大脑的电子模型中创建，同时试图尽可能多地复制其生物学细节。基于超级计算机的模拟为理解大脑的结构和功能提供了新的视角。<br />
<br />
The European Union funded a series of projects at the University of Heidelberg, which led to the development of <br />
[[BrainScaleS]] (brain-inspired multiscale computation in neuromorphic hybrid systems), a hybrid analog [[neuromorphic]] supercomputer located at Heidelberg University, Germany. It was developed as part of the [[Human Brain Project]] neuromorphic computing platform and is the complement to the [[SpiNNaker]] supercomputer (which is based on digital technology). The architecture used in BrainScaleS mimics biological neurons and their connections on a physical level; additionally, since the components are made of silicon, these model neurons operate on average 864 times (24 hours of real time is 100 seconds in the machine simulation) that of their biological counterparts.<ref>{{Cite web|date=2016-03-21|title=Beyond von Neumann, Neuromorphic Computing Steadily Advances|url=https://www.hpcwire.com/2016/03/21/lacking-breakthrough-neuromorphic-computing-steadily-advance/|access-date=2021-10-08|website=HPCwire|language=en-US}}</ref><br />
<br />
The European Union funded a series of projects at the University of Heidelberg, which led to the development of <br />
BrainScaleS (brain-inspired multiscale computation in neuromorphic hybrid systems), a hybrid analog neuromorphic supercomputer located at Heidelberg University, Germany. It was developed as part of the Human Brain Project neuromorphic computing platform and is the complement to the SpiNNaker supercomputer (which is based on digital technology). The architecture used in BrainScaleS mimics biological neurons and their connections on a physical level; additionally, since the components are made of silicon, these model neurons operate on average 864 times (24 hours of real time is 100 seconds in the machine simulation) that of their biological counterparts.<br />
<br />
欧盟资助了海德堡大学的一系列项目，这些项目导致了 BrainScaleS (神经形态混合系统中受大脑启发的多尺度计算)的发展，这是一台位于德国海德堡大学的混合模拟神经形态超级计算机。它是作为人脑计划神经形态计算平台的一部分而开发的，是 SpiNNaker 超级计算机(基于数字技术)的补充。大脑尺度中使用的体系结构模拟了生物神经元及其在物理层面上的连接; 此外，由于这些组件是由硅制成的，这些模型神经元平均运行864倍(在机器模拟中，24小时的实时时间是100秒)。<br />
<br />
===Neuromorphic sensors===<br />
The concept of neuromorphic systems can be extended to sensors (not just to computation). An example of this applied to detecting [[light]] is the [[retinomorphic sensor]] or, when employed in an array, the [[event camera]].<br />
<br />
The concept of neuromorphic systems can be extended to sensors (not just to computation). An example of this applied to detecting light is the retinomorphic sensor or, when employed in an array, the event camera.<br />
<br />
= = = 神经形态传感器 = = = 神经形态系统的概念可以扩展到传感器(而不仅仅是计算)。用于检测光线的一个例子是视网膜变形传感器，或者在阵列中使用的事件摄像机。<br />
<br />
==Ethical considerations ==<br />
While the interdisciplinary concept of neuromorphic engineering is relatively new, many of the same ethical considerations apply to neuromorphic systems as apply to [[human-like machines]] and [[artificial intelligence]] in general. However, the fact that neuromorphic systems are designed to mimic a [[human brain]] gives rise to unique ethical questions surrounding their usage.<br />
<br />
While the interdisciplinary concept of neuromorphic engineering is relatively new, many of the same ethical considerations apply to neuromorphic systems as apply to human-like machines and artificial intelligence in general. However, the fact that neuromorphic systems are designed to mimic a human brain gives rise to unique ethical questions surrounding their usage.<br />
<br />
虽然神经形态工程的跨学科概念相对较新，但许多同样的伦理考虑适用于神经形态系统，就像适用于类人机器和一般人工智能一样。然而，神经形态系统是为了模仿人类大脑而设计的这一事实引起了围绕其使用的独特的伦理问题。<br />
<br />
However, the practical debate is that neuromorphic hardware as well as artificial "neural networks" are immensely simplified models of how the brain operates or processes information at a much lower [[complex system|complexity in terms of size and functional technology]] and a much more regular structure in terms of [[brain connectivity|connectivity]]. Comparing [[neuromorphic chip]]s to the brain is a very crude comparison similar to comparing a plane to a bird just because they both have wings and a tail. The fact is that neural cognitive systems are many orders of magnitude more [[Energy efficiency (physics)|energy-]] and compute-efficient than current state-of-the-art AI and neuromorphic engineering is an attempt to narrow this gap by inspiring from the brain's mechanism just like many engineering designs have [[bioengineering|bio-inspired features]].<br />
<br />
However, the practical debate is that neuromorphic hardware as well as artificial "neural networks" are immensely simplified models of how the brain operates or processes information at a much lower complexity in terms of size and functional technology and a much more regular structure in terms of connectivity. Comparing neuromorphic chips to the brain is a very crude comparison similar to comparing a plane to a bird just because they both have wings and a tail. The fact is that neural cognitive systems are many orders of magnitude more energy- and compute-efficient than current state-of-the-art AI and neuromorphic engineering is an attempt to narrow this gap by inspiring from the brain's mechanism just like many engineering designs have bio-inspired features.<br />
<br />
然而，实际上的争论是，神经形态硬件和人工“神经网络”是大脑如何运作或处理信息的极其简化的模型，在大小和功能技术方面的复杂性要低得多，在连接方面的结构则更加规则。将神经形态芯片与大脑进行比较是一种非常粗糙的比较，类似于仅仅因为一架飞机有翅膀和一条尾巴就将它与一只鸟进行比较。事实上，神经认知系统比当前最先进的人工智能具有更多的能量和计算效率，而神经形态工程是一种通过从大脑机制中激发灵感来缩小这种差距的尝试，就像许多工程设计具有生物启发的特征一样。数量级。<br />
<br />
=== Democratic concerns===<br />
Significant ethical limitations may be placed on neuromorphic engineering due to public perception.<ref>{{Cite report|url=https://ai100.stanford.edu/sites/g/files/sbiybj9861/f/ai_100_report_0831fnl.pdf|title=Artificial Intelligence and Life in 2030|author=2015 Study Panel|date=September 2016|work=One Hundred Year Study on Artificial Intelligence (AI100)|publisher=Stanford University}}</ref> Special [[Eurobarometer]] 382: Public Attitudes Towards Robots, a survey conducted by the European Commission, found that 60% of [[European Union]] citizens wanted a ban of robots in the care of children, the elderly, or the disabled. Furthermore, 34% were in favor of a ban on robots in education, 27% in healthcare, and 20% in leisure. The European Commission classifies these areas as notably “human.” The report cites increased public concern with robots that are able to mimic or replicate human functions. Neuromorphic engineering, by definition, is designed to replicate the function of the human brain.<ref name=":1">{{Cite web|url=http://ec.europa.eu/commfrontoffice/publicopinion/archives/ebs/ebs_382_en.pdf|title=Special Eurobarometer 382: Public Attitudes Towards Robots|last=European Commission|date=September 2012|website=European Commission}}</ref><br />
<br />
Significant ethical limitations may be placed on neuromorphic engineering due to public perception. Special Eurobarometer 382: Public Attitudes Towards Robots, a survey conducted by the European Commission, found that 60% of European Union citizens wanted a ban of robots in the care of children, the elderly, or the disabled. Furthermore, 34% were in favor of a ban on robots in education, 27% in healthcare, and 20% in leisure. The European Commission classifies these areas as notably “human.” The report cites increased public concern with robots that are able to mimic or replicate human functions. Neuromorphic engineering, by definition, is designed to replicate the function of the human brain.<br />
<br />
由于公众的认知，神经形态工程学可能会受到严重的伦理限制。欧盟委员会进行的一项调查发现，60% 的欧盟公民希望禁止照顾儿童、老人或残疾人的机器人。此外，34% 的人支持禁止机器人用于教育，27% 的人支持医疗保健，20% 的人支持休闲。欧盟委员会将这些地区明显归类为“人类”报告指出，公众越来越关注能够模仿或复制人类功能的机器人。神经形态工程，顾名思义，是为了复制人脑的功能而设计的。<br />
<br />
The democratic concerns surrounding neuromorphic engineering are likely to become even more profound in the future. The European Commission found that EU citizens between the ages of 15 and 24 are more likely to think of robots as human-like (as opposed to instrument-like) than EU citizens over the age of 55. When presented an image of a robot that had been defined as human-like, 75% of EU citizens aged 15–24 said it corresponded with the idea they had of robots while only 57% of EU citizens over the age of 55 responded the same way. The human-like nature of neuromorphic systems, therefore, could place them in the categories of robots many EU citizens would like to see banned in the future.<ref name=":1" /><br />
<br />
The democratic concerns surrounding neuromorphic engineering are likely to become even more profound in the future. The European Commission found that EU citizens between the ages of 15 and 24 are more likely to think of robots as human-like (as opposed to instrument-like) than EU citizens over the age of 55. When presented an image of a robot that had been defined as human-like, 75% of EU citizens aged 15–24 said it corresponded with the idea they had of robots while only 57% of EU citizens over the age of 55 responded the same way. The human-like nature of neuromorphic systems, therefore, could place them in the categories of robots many EU citizens would like to see banned in the future.<br />
<br />
围绕神经形态工程的民主关注可能在未来变得更加深刻。欧盟委员会(European Commission)发现，15至24岁的欧盟公民比55岁以上的欧盟公民更有可能认为机器人像人(而不是像仪器)。当看到一张被定义为“类人”的机器人图片时，年龄在15岁至24岁之间的欧盟公民中有75% 的人表示，这与他们对机器人的想法相符，而55岁以上的欧盟公民中只有57% 的人有同样的反应。因此，类似人类的神经形态系统，可以把它们归入许多欧盟公民希望在未来禁止使用的机器人类别。<br />
<br />
===Personhood ===<br />
As neuromorphic systems have become increasingly advanced, some scholars{{who|date=August 2021}} have advocated for granting [[personhood]] rights to these systems. If the brain is what grants humans their personhood, to what extent does a neuromorphic system have to mimic the human brain to be granted personhood rights? Critics of technology development in the [[Human Brain Project]], which aims to advance brain-inspired computing, have argued that advancement in neuromorphic computing could lead to machine consciousness or personhood.<ref>{{Cite journal|last=Aicardi|first=Christine|date=September 2018|title=Accompanying technology development in the Human Brain Project: From foresight to ethics management|journal=Futures|volume=102|pages=114–124|doi=10.1016/j.futures.2018.01.005|doi-access=free}}</ref> If these systems are to be treated as people, critics argue, then many tasks humans perform using neuromorphic systems, including the act of termination of neuromorphic systems, may be morally impermissible as these acts would violate the autonomy of the neuromorphic systems.<ref>{{Cite journal|last=Lim|first=Daniel|date=2014-06-01|title=Brain simulation and personhood: a concern with the Human Brain Project|journal=Ethics and Information Technology|language=en|volume=16|issue=2|pages=77–89|doi=10.1007/s10676-013-9330-5|s2cid=17415814|issn=1572-8439}}</ref><br />
<br />
As neuromorphic systems have become increasingly advanced, some scholars have advocated for granting personhood rights to these systems. If the brain is what grants humans their personhood, to what extent does a neuromorphic system have to mimic the human brain to be granted personhood rights? Critics of technology development in the Human Brain Project, which aims to advance brain-inspired computing, have argued that advancement in neuromorphic computing could lead to machine consciousness or personhood. If these systems are to be treated as people, critics argue, then many tasks humans perform using neuromorphic systems, including the act of termination of neuromorphic systems, may be morally impermissible as these acts would violate the autonomy of the neuromorphic systems.<br />
<br />
= = = 人格 = = 随着神经形态系统的日益发展，一些学者主张赋予这些系统人格权。如果是大脑赋予了人类人格，那么神经形态系统在多大程度上必须模仿人类大脑才能被赋予人格权利？“人脑计划”旨在推进以大脑为灵感的计算机技术发展，该计划的批评者认为，神经形态计算机技术的进步可能导致机器意识或人格的形成。批评家认为，如果这些系统被当作人来对待，那么人类使用神经形态系统执行的许多任务，包括终止神经形态系统的行为，可能在道德上是不允许的，因为这些行为将违反神经形态系统的自主性。<br />
<br />
==Dual use (military applications)==<br />
The [[Joint Artificial Intelligence Center]], a branch of the U.S. military, is a center dedicated to the procurement and implementation of AI software and neuromorphic hardware for combat use. Specific applications include smart headsets/goggles and robots. JAIC intends to rely heavily on neuromorphic technology to connect "every fighter every shooter" within a network of neuromorphic-enabled units.<br />
<br />
The Joint Artificial Intelligence Center, a branch of the U.S. military, is a center dedicated to the procurement and implementation of AI software and neuromorphic hardware for combat use. Specific applications include smart headsets/goggles and robots. JAIC intends to rely heavily on neuromorphic technology to connect "every fighter every shooter" within a network of neuromorphic-enabled units.<br />
<br />
军民两用联合人工智能中心是美国军队的一个分支，专门从事采购和实施用于战斗的人工智能软件和神经形态硬件。具体应用包括智能耳机/护目镜和机器人。JAIC 打算严重依赖神经形态技术来连接神经形态单位网络中的“每个战士每个射手”。<br />
<br />
==Legal considerations==<br />
Skeptics have argued that there is no way to apply the electronic personhood, the concept of personhood that would apply to neuromorphic technology, legally. In a letter signed by 285 experts in law, robotics, medicine, and ethics opposing a European Commission proposal to recognize “smart robots” as legal persons, the authors write, “A legal status for a robot can’t derive from the [[Natural person|Natural Person]] model, since the robot would then hold [[human rights]], such as the right to dignity, the right to its integrity, the right to remuneration or the right to citizenship, thus directly confronting the Human rights. This would be in contradiction with the [[Charter of Fundamental Rights of the European Union]] and the [[Convention for the Protection of Human Rights and Fundamental Freedoms]].”<ref>{{Cite web|url=http://www.robotics-openletter.eu/|title=Robotics Openletter {{!}} Open letter to the European Commission|language=fr-FR|access-date=2019-05-10}}</ref><br />
<br />
Skeptics have argued that there is no way to apply the electronic personhood, the concept of personhood that would apply to neuromorphic technology, legally. In a letter signed by 285 experts in law, robotics, medicine, and ethics opposing a European Commission proposal to recognize “smart robots” as legal persons, the authors write, “A legal status for a robot can’t derive from the Natural Person model, since the robot would then hold human rights, such as the right to dignity, the right to its integrity, the right to remuneration or the right to citizenship, thus directly confronting the Human rights. This would be in contradiction with the Charter of Fundamental Rights of the European Union and the Convention for the Protection of Human Rights and Fundamental Freedoms.”<br />
<br />
= = 法律方面的考虑 = = 怀疑论者认为，在法律上没有办法应用电子人格，这个人格概念将适用于神经形态技术。在一封由285名法律、机器人、医学和伦理学专家签名的信中，作者们反对欧盟委员会承认“智能机器人”为法人的提议，他们写道，“机器人的法律地位不能从自然人模型中推导出来，因为机器人将拥有人权，如尊严权、完整权、报酬权或公民权，从而直接面对人权。这将有悖于《欧洲联盟基本权利宪章和《保护人权和基本自由公约》。”<br />
<br />
===Ownership and property rights===<br />
There is significant legal debate around property rights and artificial intelligence. In ''Acohs Pty Ltd v. Ucorp Pty Ltd'', Justice Christopher Jessup of the [[Federal Court of Australia]] found that the [[source code]] for [[Material safety data sheets|Material Safety Data Sheets]] could not be [[Copyright law of Australia|copyrighted]] as it was generated by a [[software interface]] rather than a human author.<ref>{{Cite web|url=http://www.lavan.com.au/advice/intellectual_property_technology/copyright_in_source_code_and_digital_products|title=Copyright in source code and digital products|last=Lavan|website=Lavan|language=en|access-date=2019-05-10}}</ref> The same question may apply to neuromorphic systems: if a neuromorphic system successfully mimics a human brain and produces a piece of original work, who, if anyone, should be able to claim ownership of the work?<ref>{{cite journal |last1=Eshraghian|first1=Jason K. |title=Human Ownership of Artificial Creativity |journal=Nature Machine Intelligence |date=9 March 2020 |volume=2 |pages=157–160 |doi=10.1038/s42256-020-0161-x}}</ref><br />
<br />
There is significant legal debate around property rights and artificial intelligence. In Acohs Pty Ltd v. Ucorp Pty Ltd, Justice Christopher Jessup of the Federal Court of Australia found that the source code for Material Safety Data Sheets could not be copyrighted as it was generated by a software interface rather than a human author. The same question may apply to neuromorphic systems: if a neuromorphic system successfully mimics a human brain and produces a piece of original work, who, if anyone, should be able to claim ownership of the work?<br />
<br />
= = = 所有权和财产权 = = = 围绕财产权和人工智能有着重大的法律争论。在 Acohs Pty Ltd 诉 Ucorp Pty Ltd 一案中，澳大利亚联邦法院的克里斯托弗 · 杰瑟普法官发现，材料安全数据表的源代码不能受版权保护，因为它是由软件界面而不是人工作者生成的。同样的问题可能也适用于神经形态系统: 如果一个神经形态系统成功地模仿了人类的大脑并产生了一部原创作品，那么谁，如果有人，应该声称拥有这部作品的所有权？<br />
<br />
== Neuromemristive systems==<br />
Neuromemristive systems are a subclass of neuromorphic computing systems that focus on the use of [[memristors]] to implement [[neuroplasticity]]. While neuromorphic engineering focuses on mimicking biological behavior, neuromemristive systems focus on abstraction.<ref>{{Cite web|url=https://digitalops.sandia.gov/Mediasite/Play/a10cf6ceb55d47608bb8326dd00e46611d|title=002.08 N.I.C.E. Workshop 2014: Towards Intelligent Computing with Neuromemristive Circuits and Systems - Feb. 2014|website=digitalops.sandia.gov|access-date=2019-08-26}}</ref> For example, a neuromemristive system may replace the details of a [[Cerebral cortex|cortical]] microcircuit's behavior with an abstract neural network model.<ref>C. Merkel and D. Kudithipudi, "Neuromemristive extreme learning machines for pattern classification," ISVLSI, 2014.</ref><br />
<br />
Neuromemristive systems are a subclass of neuromorphic computing systems that focus on the use of memristors to implement neuroplasticity. While neuromorphic engineering focuses on mimicking biological behavior, neuromemristive systems focus on abstraction. For example, a neuromemristive system may replace the details of a cortical microcircuit's behavior with an abstract neural network model.C. Merkel and D. Kudithipudi, "Neuromemristive extreme learning machines for pattern classification," ISVLSI, 2014.<br />
<br />
神经记忆电阻系统是神经形态计算系统的一个亚类，主要研究利用记忆电阻器实现神经可塑性。神经形态工程的重点是模拟生物行为，而神经记忆电阻系统的重点是提取。例如，一个神经记忆系统可以用一个抽象的神经网络模型取代皮层微电路的行为细节。默克尔和 d. Kudithipudi，“用于模式分类的神经记忆极端学习机器”，ISVLSI，2014。<br />
<br />
There exist several neuron inspired threshold logic functions<ref name="Maan 1–13" /> implemented with memristors that have applications in high level [[pattern recognition]] applications. Some of the applications reported recently include [[speech recognition]],<ref>{{Cite journal|title = Memristor pattern recogniser: isolated speech word recognition|journal = Electronics Letters|pages = 1370–1372|volume = 51|issue = 17|doi = 10.1049/el.2015.1428|first1 = A.K.|last1 = Maan|first2 = A.P.|last2 = James|first3 = S.|last3 = Dimitrijev|year = 2015|bibcode = 2015ElL....51.1370M|hdl = 10072/140989|s2cid = 61454815|url = https://semanticscholar.org/paper/48d3ab11ec6e213b62f11eedcfb7b7febb058674|hdl-access = free}}</ref> [[face recognition]]<ref>{{Cite journal|title = Memristive Threshold Logic Face Recognition|journal = Procedia Computer Science|date = 2014-01-01|pages = 98–103|volume = 41|series = 5th Annual International Conference on Biologically Inspired Cognitive Architectures, 2014 BICA|doi = 10.1016/j.procs.2014.11.090|first1 = Akshay Kumar|last1 = Maan|first2 = Dinesh S.|last2 = Kumar|first3 = Alex Pappachen|last3 = James|doi-access = free}}</ref> and [[object recognition]].<ref>{{Cite journal|title = Memristive Threshold Logic Circuit Design of Fast Moving Object Detection|journal = IEEE Transactions on Very Large Scale Integration (VLSI) Systems|date = 2015-10-01|issn = 1063-8210|pages = 2337–2341|volume = 23|issue = 10|doi = 10.1109/TVLSI.2014.2359801|first1 = A.K.|last1 = Maan|first2 = D.S.|last2 = Kumar|first3 = S.|last3 = Sugathan|first4 = A.P.|last4 = James|arxiv = 1410.1267|s2cid = 9647290}}</ref> They also find applications in replacing conventional digital logic gates.<ref>{{Cite journal|title = Resistive Threshold Logic|journal = IEEE Transactions on Very Large Scale Integration (VLSI) Systems|date = 2014-01-01|issn = 1063-8210|pages = 190–195|volume = 22|issue = 1|doi = 10.1109/TVLSI.2012.2232946|first1 = A.P.|last1 = James|first2 = L.R.V.J.|last2 = Francis|first3 = D.S.|last3 = Kumar|arxiv = 1308.0090|s2cid = 7357110}}</ref><ref>{{Cite journal|title = Threshold Logic Computing: Memristive-CMOS Circuits for Fast Fourier Transform and Vedic Multiplication|journal = IEEE Transactions on Very Large Scale Integration (VLSI) Systems|date = 2015-11-01|issn = 1063-8210|pages = 2690–2694|volume = 23|issue = 11|doi = 10.1109/TVLSI.2014.2371857|first1 = A.P.|last1 = James|first2 = D.S.|last2 = Kumar|first3 = A.|last3 = Ajayan|arxiv = 1411.5255|s2cid = 6076956}}</ref><br />
<br />
There exist several neuron inspired threshold logic functions implemented with memristors that have applications in high level pattern recognition applications. Some of the applications reported recently include speech recognition, face recognition and object recognition. They also find applications in replacing conventional digital logic gates.<br />
<br />
利用记忆电阻器实现的神经元启发阈值逻辑函数在高级模式识别中有着广泛的应用。最近报道的一些应用包括语音识别、人脸识别和物体识别。它们还可以用来取代传统的数字逻辑门。<br />
<br />
For ideal passive memristive circuits there is an exact equation (Caravelli-Traversa-[[Di Ventra]] equation) for the internal memory of the circuit:<ref>{{cite journal |last=Caravelli |display-authors=etal|arxiv=1608.08651 |title=The complex dynamics of memristive circuits: analytical results and universal slow relaxation |year=2017 |doi=10.1103/PhysRevE.95.022140 |pmid= 28297937 |volume=95 |issue= 2 |pages= 022140 |journal=Physical Review E|bibcode=2017PhRvE..95b2140C |s2cid=6758362}}</ref><br />
<br />
For ideal passive memristive circuits there is an exact equation (Caravelli-Traversa-Di Ventra equation) for the internal memory of the circuit:<br />
<br />
对于理想的无源记忆电路，电路的内部记忆有一个精确的方程(Caravelli-Traversa-Di Ventra 方程) :<br />
<br />
:<math> \frac{d}{dt} \vec{W} = \alpha \vec{W}-\frac{1}{\beta} (I+\xi \Omega W)^{-1} \Omega \vec S </math><br />
<br />
:\frac{d}{dt} \vec{W} = \alpha \vec{W}-\frac{1}{\beta} (I+\xi \Omega W)^{-1} \Omega \vec S<br />
<br />
: \frac{d}{dt} \vec{W} = \alpha \vec{W}-\frac{1}{\beta} (I+\xi \Omega W)^{-1} \Omega \vec S<br />
<br />
as a function of the properties of the physical memristive network and the external sources. In the equation above, <math>\alpha</math> is the "forgetting" time scale constant, <math> \xi=r-1</math> and <math>r=\frac{R_\text{off}}{R_\text{on}}</math> is the ratio of ''off'' and ''on'' values of the limit resistances of the memristors, <math> \vec S </math> is the vector of the sources of the circuit and <math>\Omega</math> is a projector on the fundamental loops of the circuit. The constant <math>\beta</math> has the dimension of a voltage and is associated to the properties of the [[memristor]]; its physical origin is the charge mobility in the conductor. The diagonal matrix and vector <math>W=\operatorname{diag}(\vec W)</math> and <math>\vec W</math> respectively, are instead the internal value of the memristors, with values between 0 and 1. This equation thus requires adding extra constraints on the memory values in order to be reliable.<br />
<br />
<nowiki>as a function of the properties of the physical memristive network and the external sources. In the equation above, \alpha is the "forgetting" time scale constant, \xi=r-1 and r=\frac{R_\text{off}}{R_\text{on}} is the ratio of off and on values of the limit resistances of the memristors, \vec S is the vector of the sources of the circuit and \Omega is a projector on the fundamental loops of the circuit. The constant \beta has the dimension of a voltage and is associated to the properties of the memristor; its physical origin is the charge mobility in the conductor. The diagonal matrix and vector W=\operatorname{diag}(\vec W) and \vec W respectively, are instead the internal value of the memristors, with values between 0 and 1. This equation thus requires adding extra constraints on the memory values in order to be reliable.</nowiki><br />
<br />
<nowiki>作为物理记忆网络和外部源的性质的函数。在上述方程中，α 是“遗忘”时间尺度常数，xi = r-1，r = frac { r _ text { off }{ on }{ r _ text { on }}是记忆电阻器极限电阻的开关和开关值之比，vec s 是电路源的矢量，Omega 是电路基本环路的投影仪。常数 β 具有电压的尺寸，与记忆电阻器的特性有关; 它的物理起源是导体中的电荷迁移率。对角矩阵和向量 w = 操作者名{ diag }(vec w)和 vec w 分别是记忆电阻器的内值，值在0到1之间。因此，这个等式需要在内存值上添加额外的约束，以保证可靠性。</nowiki><br />
<br />
== See also==<br />
{{Columns-list|colwidth=18em|<br />
* [[AI accelerator (computer hardware)|AI accelerator]]<br />
* [[Artificial brain]]<br />
* [[Biomorphic]]<br />
* [[Cognitive computer]]<br />
* [[Computation and Neural Systems]]<br />
* [[Differentiable programming]]<br />
* [[Event camera]]<br />
* [[Neurorobotics]]<br />
* [[Optical flow sensor]]<br />
* [[Physical neural network]]<br />
* [[SpiNNaker]]<br />
* [[SyNAPSE]]<br />
* [[Retinomorphic sensor]]<br />
* [[Vision chip]]<br />
* [[Vision processing unit]]<br />
* [[Zeroth (software)]]<br />
* [[Hardware for artificial intelligence]]<br />
}}<br />
<br />
{{Portal bar|Electronics}}<br />
<br />
==References==<br />
{{Reflist|40em}}<br />
<br />
==External links==<br />
<!--======================== {{No more links}} ============================<br />
| PLEASE BE CAUTIOUS IN ADDING MORE LINKS TO THIS ARTICLE. Wikipedia |<br />
| is not a collection of links nor should it be used for advertising. |<br />
| |<br />
| Excessive or inappropriate links WILL BE DELETED. |<br />
| See [[Wikipedia:External links]] & [[Wikipedia:Spam]] for details. |<br />
| |<br />
| If there are already plentiful links, please propose additions or |<br />
| replacements on this article's discussion page, or submit your link |<br />
| to the relevant category at the Open Directory Project (dmoz.org) |<br />
| and link back to that category using the {{dmoz}} template. |<br />
======================={{No more links}}=============================--><br />
{{Commons category|Neuromorphic Engineering}}<br />
*[https://web.archive.org/web/20150727034331/http://ine-web.org/workshops/workshops-overview Telluride Neuromorphic Engineering Workshop]<br />
*[https://archive.today/20130115190057/http://capocaccia.ethz.ch/ CapoCaccia Cognitive Neuromorphic Engineering Workshop]<br />
*[https://web.archive.org/web/20190716132350/http://www.ine-web.org/ Institute of Neuromorphic Engineering]<br />
*[http://www.ine-news.org/ INE news site].<br />
*[http://www.frontiersin.org/neuromorphic_engineering Frontiers in Neuromorphic Engineering Journal]<br />
*[http://www.cns.caltech.edu/ Computation and Neural Systems] department at the [[California Institute of Technology]].<br />
*[http://www.humanbrainproject.eu/ Human Brain Project official site]<br />
*[https://www.the-scientist.com/features/building-a-silicon-brain-65738 Building a Silicon Brain:] Computer chips based on biological neurons may help simulate larger and more-complex brain models. May 1, 2019. SANDEEP RAVINDRAN<br />
<br />
<br />
*Telluride Neuromorphic Engineering Workshop<br />
*CapoCaccia Cognitive Neuromorphic Engineering Workshop<br />
*Institute of Neuromorphic Engineering<br />
*INE news site.<br />
*Frontiers in Neuromorphic Engineering Journal<br />
*Computation and Neural Systems department at the California Institute of Technology.<br />
*Human Brain Project official site<br />
*Building a Silicon Brain: Computer chips based on biological neurons may help simulate larger and more-complex brain models. May 1, 2019. SANDEEP RAVINDRAN<br />
<br />
=<nowiki>= = 外部链接 =</nowiki>= <br />
*碲化物神经形态工程工作室<br />
*CapoCaccia 认知神经形态工程工作室<br />
*神经形态工程研究所<br />
*INE 新闻站点。<br />
*《神经形态工程学前沿》<br />
*加州理工学院计算与神经系统系。<br />
* 人脑项目官方网站<br />
*打造硅脑: 基于生物神经元的计算机芯片可能有助于模拟更大、更复杂的大脑模型。2019年5月1日。SANDEEP RAVINDRAN<br />
<br />
{{Differentiable computing}}<br />
{{Authority control}}<br />
<br />
[[Category:Electrical engineering]]<br />
[[Category:Neuroscience]]<br />
[[分类:AI accelerators|*]]<br />
[[Category:Artificial intelligence]]<br />
[[Category:Robotics]]<br />
<br />
Category:Electrical engineering<br />
Category:Neuroscience<br />
*<br />
Category:Artificial intelligence<br />
Category:Robotics<br />
<br />
类别: 电气工程类别: 神经科学 * 类别: 人工智能类别: 机器人<br />
<br />
<noinclude><br />
<br />
<small>This page was moved from [[wikipedia:en:Neuromorphic engineering]]. Its edit history can be viewed at [[神经形态计算/edithistory]]</small></noinclude><br />
<br />
[[Category:待整理页面]]<br />
<br />
<references /></div>Shenky20https://wiki.swarma.org/index.php?title=%E5%A4%A7%E8%A7%84%E6%A8%A1%E8%84%91%E7%BD%91%E7%BB%9C&diff=29995大规模脑网络2022-04-04T09:55:01Z<p>Shenky20：</p>
<hr />
<div>此词条由神经动力学模型读书会词条梳理志愿者Shenky20翻译审校，翻译字数共800，未经专家审核，带来阅读不便，请见谅。<br />
<br />
{{Short description|Collections of brain regions}}<br />
'''Large-scale brain networks''' (also known as '''intrinsic brain networks''') are collections of widespread [[brain regions]] showing [[Resting state fMRI#Functional|functional connectivity]] by statistical analysis of the [[Functional magnetic resonance imaging|fMRI]] [[BOLD signal]]<ref name=Riedl>{{cite journal|last1=Riedl|first1=Valentin|last2=Utz|first2=Lukas|last3=Castrillón|first3=Gabriel|last4=Grimmer|first4=Timo|last5=Rauschecker|first5=Josef P.|last6=Ploner|first6=Markus|last7=Friston|first7=Karl J.|last8=Drzezga|first8=Alexander|last9=Sorg|first9=Christian|title=Metabolic connectivity mapping reveals effective connectivity in the resting human brain|journal=PNAS|date=January 12, 2016|volume=113|issue=2|pages=428–433|doi=10.1073/pnas.1513752113|pmid=26712010|pmc=4720331|bibcode=2016PNAS..113..428R|doi-access=free}}</ref> or other recording methods such as [[Electroencephalography|EEG]],<ref name=":1">{{Cite journal|last1=Foster|first1=Brett L.|last2=Parvizi|first2=Josef|date=2012-03-01|title=Resting oscillations and cross-frequency coupling in the human posteromedial cortex|journal=NeuroImage|volume=60|issue=1|pages=384–391|doi=10.1016/j.neuroimage.2011.12.019|pmid=22227048|issn=1053-8119|pmc=3596417}}</ref> [[Positron emission tomography|PET]]<ref name=":2">{{Cite journal|last1=Buckner|first1=Randy L.|last2=Andrews‐Hanna|first2=Jessica R.|last3=Schacter|first3=Daniel L.|date=2008|title=The Brain's Default Network|journal=Annals of the New York Academy of Sciences|language=en|volume=1124|issue=1|pages=1–38|doi=10.1196/annals.1440.011|pmid=18400922|issn=1749-6632|bibcode=2008NYASA1124....1B|s2cid=3167595}}</ref> and [[Magnetoencephalography|MEG]].<ref name=":3">{{Cite journal|last1=Morris|first1=Peter G.|last2=Smith|first2=Stephen M.|last3=Barnes|first3=Gareth R.|last4=Stephenson|first4=Mary C.|last5=Hale|first5=Joanne R.|last6=Price|first6=Darren|last7=Luckhoo|first7=Henry|last8=Woolrich|first8=Mark|last9=Brookes|first9=Matthew J.|date=2011-10-04|title=Investigating the electrophysiological basis of resting state networks using magnetoencephalography|journal=Proceedings of the National Academy of Sciences|language=en|volume=108|issue=40|pages=16783–16788|doi=10.1073/pnas.1112685108|issn=0027-8424|pmid=21930901|pmc=3189080|bibcode=2011PNAS..10816783B|doi-access=free}}</ref> An emerging paradigm in neuroscience is that cognitive tasks are performed not by individual brain regions working in isolation but by networks consisting of several discrete brain regions that are said to be "functionally connected". Functional connectivity networks may be found using algorithms such as [[cluster analysis]], spatial [[independent component analysis]] (ICA), seed based, and others.<ref name="Petersen">{{cite journal|last1=Petersen|first1=Steven|last2=Sporns|first2=Olaf|title=Brain Networks and Cognitive Architectures|journal=Neuron|date=October 2015|volume=88|issue=1|pages=207–219|doi=10.1016/j.neuron.2015.09.027|pmid=26447582|pmc=4598639 }}</ref> Synchronized brain regions may also be identified using long-range synchronization of the EEG, MEG, or other dynamic brain signals.<ref name=Bressler>{{cite journal|last1=Bressler|first1=Steven L.|last2=Menon|first2=Vinod|s2cid=5967761|title=Large scale brain networks in cognition: emerging methods and principles|journal=Trends in Cognitive Sciences|date=June 2010|volume=14|issue=6|pages=233–290|doi=10.1016/j.tics.2010.04.004|url=http://www.cell.com/trends/cognitive-sciences/issue?pii=S1364-6613(10)X0005-5|accessdate=24 January 2016|pmid=20493761}}</ref><br />
<br />
'''<font color="#ff8000">大规模脑网络Large-scale brain networks</font>'''(也称为'''<font color="#ff8000">内在大脑网络Intrinsic brain networks</font>''')是在对基于'''<font color="#ff8000">血氧水平依赖效应BOLD</font>'''的'''<font color="#ff8000">功能性磁共振成像fMRI</font>'''信号<ref name="Riedl" />的统计分析或其他记录方法(如'''<font color="#ff8000">脑电图EEG<ref name=":1" /></font>'''、'''<font color="#ff8000">正电子发射断层扫描技术PET<ref name=":2" /></font>'''和'''<font color="#ff8000">脑磁图MEG<ref name=":3" /></font>''')中，表现出'''<font color="#ff8000">功能连接Functional connectivity</font>'''的'''<font color="#ff8000">脑区Brain regions</font>'''的集合。根据神经科学中一个新出现的范式，认知任务不是由单个脑区独立执行的，而是由几个互不相连的脑区“功能连接”组成的网络执行的。功能连接网络可以通过'''<font color="#ff8000">数据聚类Cluster analysis</font>'''、空间'''<font color="#ff8000">独立元素分析ICA</font>'''、种子点方法等算法来发现。<ref name="Petersen" />同步的脑区也可以用脑电图、脑磁图或其他动态脑信号的远程同步来识别。<ref name="Bressler" /><br />
<br />
The set of identified brain areas that are linked together in a large-scale network varies with cognitive function.<ref name="Bressler2">{{cite journal|last1=Bressler|first1=Steven L.|title=Neurocognitive networks|journal=Scholarpedia|volume=3|issue=2|pages=1567|doi=10.4249/scholarpedia.1567|year=2008|bibcode=2008SchpJ...3.1567B|doi-access=free}}</ref> When the cognitive state is not explicit (i.e., the subject is at "rest"), the large-scale brain network is a [[Resting state fMRI|resting state]] network (RSN). As a physical system with graph-like properties,<ref name="Bressler" /> a large-scale brain network has both nodes and edges and cannot be identified simply by the co-activation of brain areas. In recent decades, the analysis of brain networks was made feasible by advances in imaging techniques as well as new tools from [[graph theory]] and [[Dynamical systems theory|dynamical systems]].<br />
<br />
<br />
大规模脑网络中,连接在一起的脑区集合因认知功能的不同而不同。<ref name="Bressler2" />当认知状态不明确(即主体处于“静止”状态)时，大规模脑网络是一个'''<font color="#ff8000">静息状态Resting state</font>'''网络(RSN)。作为一个具有图形特征的物理系统，<ref name="Bressler" />大规模脑网络既有节点又有边，不能简单地通过脑区的共同激活来识别。近几十年来，成像技术不断进步，同时'''<font color="#ff8000">图论Graph theory</font>'''、'''<font color="#ff8000">动力学系统Dynamical systems</font>'''领域出现了新的技术手段，这使得脑网络分析变得可行。<br />
<br />
Large-scale brain networks are identified by their function and provide a coherent framework for understanding [[cognition]] by offering a neural model of how different cognitive functions emerge when different sets of brain regions join together as self-organized coalitions. The number and composition of the coalitions will vary with the algorithm and parameters used to identify them.<ref name="Yeo" /><ref name=":4">{{cite journal|last1=Abou Elseoud|first1=Ahmed|last2=Littow|first2=Harri|last3=Remes|first3=Jukka|last4=Starck|first4=Tuomo|last5=Nikkinen|first5=Juha|last6=Nissilä|first6=Juuso|last7=Timonen|first7=Markku|last8=Tervonen|first8=Osmo|last9=Kiviniemi1|first9=Vesa|title=Group-ICA Model Order Highlights Patterns of Functional Brain Connectivity|journal= Frontiers in Systems Neuroscience|date=2011-06-03|volume=5|pages=37|doi=10.3389/fnsys.2011.00037|pmid=21687724|pmc=3109774 |doi-access=free}}</ref> In one model, there is only the [[default mode network]] and the [[task-positive network]], but most current analyses show several networks, from a small handful to 17.<ref name="Yeo" /> The most common and stable networks are enumerated below. The regions participating in a functional network may be dynamically reconfigured.<ref name="Petersen" /><ref name="Bassett">{{cite journal|last1=Bassett|first1=Daniella|last2=Bertolero|first2=Max|title=How Matter Becomes Mind|journal=Scientific American|date=July 2019|volume=321|issue=1|page=32|url=https://www.scientificamerican.com/|accessdate=23 June 2019 }}</ref><br />
<br />
<br />
大规模脑网络是通过其功能来进行识别的。通过研究大规模脑网络建立神经模型，对不同脑区组合所形成的自组织联合体如何实现不同的'''<font color="#ff8000">认知Cognition</font>'''功能进行解释，就能够为认知理解提供一个连贯的框架。识别算法和参数的不同会导致所识别出的上述联合体的数量和组成有所不同。<ref name="Yeo" /><ref name=":4" />一个模型理论认为，符合上述条件的神经模型只包含'''<font color="#ff8000">默认模式网络Default mode network</font>'''和'''<font color="#ff8000">任务激活网络Task-positive network</font>'''，但目前'''<font color="#32CD32">大多数分析理论都包括从几个到17个不等的网络。<ref name="Yeo" /></font>'''下面列举了最常见且稳定的网络。'''<font color="#32CD32"> 人脑</font>'''可以动态地重新配置参与功能网络的脑区。<ref name="Petersen" /><ref name="Bassett" /> <br />
<br />
Disruptions in activity in various networks have been implicated in neuropsychiatric disorders such as [[Depression (mood)|depression]], [[Alzheimer's disease|Alzheimer's]], [[Autism-spectrum disorder|autism spectrum disorder]], [[schizophrenia]], [[ADHD]]<ref name=":5">{{cite journal |last1=Griffiths |first1=Kristi R. |last2=Braund |first2=Taylor A. |last3=Kohn |first3=Michael R. |last4=Clarke |first4=Simon |last5=Williams |first5=Leanne M. |last6=Korgaonkar |first6=Mayuresh S. |title=Structural brain network topology underpinning ADHD and response to methylphenidate treatment |journal=Translational Psychiatry |date=2 March 2021 |volume=11 |issue=1 |pages=1–9 |doi=10.1038/s41398-021-01278-x | pmc=7925571 |pmid=33654073 |url=https://www.nature.com/articles/s41398-021-01278-x#citeas |access-date=16 November 2021}}</ref> and [[bipolar disorder]].<ref name=":6">{{Cite journal|url=https://www.researchgate.net/publication/51639686|title=Large-scale brain networks and psychopathology: A unifying triple network model|last=Menon|first=Vinod|s2cid=26653572|journal=Trends in Cognitive Sciences|date=2011-09-09|volume=15|issue=10|pages=483–506|doi=10.1016/j.tics.2011.08.003|pmid=21908230}}</ref><br />
<br />
<br />
脑网络活动的中断与诸多神经精神疾病密切相关，如'''<font color="#ff8000">抑郁症Depression</font>'''、'''<font color="#ff8000">老年痴呆症Alzheimer's</font>'''、'''<font color="#ff8000">自闭症谱系障碍Autism spectrum disorder</font>'''、'''<font color="#ff8000">精神分裂症Schizophrenia</font>'''、'''<font color="#ff8000">多动症ADHD<ref name=":5" /></font>'''和'''<font color="#ff8000">躁郁症Bipolar disorder<ref name=":6" /></font>'''。<br />
<br />
==Core networks==<br />
[[File:Heine2012x3010.png|thumb|An example that identified 10 large-scale brain networks from [[resting state fMRI]] activity through [[independent component analysis]].<ref name="Heine" />|链接=Special:FilePath/Heine2012x3010.png]]<br />
Because brain networks can be identified at various different resolutions and with various different neurobiological properties, there is no such thing as a universal atlas of brain networks that fits all circumstances.<ref name=":7">{{cite journal|last1=Eickhoff|first1=SB|last2=Yeo|first2=BTT|last3=Genon|first3=S|title=Imaging-based parcellations of the human brain.|journal=Nature Reviews. Neuroscience|date=November 2018|volume=19|issue=11|pages=672–686|doi=10.1038/s41583-018-0071-7|pmid=30305712|s2cid=52954265|url=http://juser.fz-juelich.de/record/856633/files/Eickhoff_Yeo_Genon_NRN_MainManuscriptInclFigures.pdf}}</ref> While acknowledging this problem, Uddin, Yeo, and Spreng proposed in 2019<ref name="Uddin2019">{{cite journal|last1=Uddin|first1=LQ|last2=Yeo|first2=BTT|last3=Spreng|first3=RN|title=Towards a Universal Taxonomy of Macro-scale Functional Human Brain Networks.|journal=Brain Topography|date=November 2019|volume=32|issue=6|pages=926–942|doi=10.1007/s10548-019-00744-6|pmid=31707621|pmc=7325607}}</ref> that the following six networks should be defined as core networks based on converging evidences from multiple studies<ref name=":8">{{cite journal|last1=Doucet|first1=GE|last2=Lee|first2=WH|last3=Frangou|first3=S|title=Evaluation of the spatial variability in the major resting-state networks across human brain functional atlases.|journal=Human Brain Mapping|date=2019-10-15|volume=40|issue=15|pages=4577–4587|doi=10.1002/hbm.24722|pmid=31322303|pmc=6771873}}</ref><ref name="Yeo" /><ref name=":9">{{cite journal|last1=Smith|first1=SM|last2=Fox|first2=PT|last3=Miller|first3=KL|last4=Glahn|first4=DC|last5=Fox|first5=PM|last6=Mackay|first6=CE|last7=Filippini|first7=N|last8=Watkins|first8=KE|last9=Toro|first9=R|last10=Laird|first10=AR|last11=Beckmann|first11=CF|title=Correspondence of the brain's functional architecture during activation and rest.|journal=Proceedings of the National Academy of Sciences of the United States of America|date=2009-08-04|volume=106|issue=31|pages=13040–5|doi=10.1073/pnas.0905267106|pmid=19620724|pmc=2722273|bibcode=2009PNAS..10613040S|doi-access=free}}</ref> to facilitate communication between researchers.<br />
<br />
==核心网络==<br />
【图1：An example that identified 10 large-scale brain networks from resting state fMRI activity through independent component analysis.<ref name="Heine" />这是一个通过独立元素分析方法从静息状态脑功能性磁共振成像信号中分辨出10个大规模脑网络的例子。<ref name="Heine" />】<br />
<br />
<br />
因为脑网络可以在不同分辨率条件下、使用不同的神经生物学特性来识别，所以没有能够适用于所有情况的通用脑网络图谱。<ref name=":7" />在承认这个问题的同时，Uddin，Yeo，和Spreng在2019年提出<ref name="Uddin2019" />，综合考虑来自多个研究的证据<ref name=":8" /><ref name="Yeo" /><ref name=":9" />，以下六个网络应该被定义为核心网络，以促进研究人员之间的交流。<br />
<br />
===Default Mode (Medial frontoparietal)===<br />
{{Main|Default mode network}}<br />
*The default mode network is active when an individual is awake and at rest. It preferentially activates when individuals focus on internally-oriented tasks such as daydreaming, envisioning the future, retrieving memories, and [[theory of mind]]. It is negatively correlated with brain systems that focus on external visual signals. It is the most widely researched network.<ref name="Bressler" /><ref name="Bassett" /><ref name=":10">{{Cite journal|date=2012-08-15|title=The serendipitous discovery of the brain's default network|journal=NeuroImage|language=en|volume=62|issue=2|pages=1137–1145|doi=10.1016/j.neuroimage.2011.10.035|pmid=22037421|issn=1053-8119|last1=Buckner|first1=Randy L.|s2cid=9880586}}</ref><ref name="Riedl" /><ref name="Yuan"><br />
{{cite journal|last1=Yuan|first1=Rui|last2=Di|first2=Xin|last3=Taylor|first3=Paul A.|last4=Gohel|first4=Suril|last5=Tsai|first5=Yuan-Hsiung|last6=Biswal|first6=Bharat B.|title=Functional topography of the thalamocortical system in human|journal=Brain Structure and Function|date=30 April 2015|doi=10.1007/s00429-015-1018-7|pmid=25924563|pmc=6363530|volume=221|issue=4|pages=1971–1984}}</ref><ref name="Bell">{{cite journal|last1=Bell|first1=Peter T.|last2=Shine|first2=James M.|title=Estimating Large-Scale Network Convergence in the Human Functional Connectome|journal=Brain Connectivity|date=2015-11-09|volume=5|issue=9|doi=10.1089/brain.2015.0348|pmid=26005099|pages=565–74}}</ref><ref name="Heine">{{cite journal|last1=Heine|first1=Lizette|last2=Soddu|first2=Andrea|last3=Gomez|first3=Francisco|last4=Vanhaudenhuyse|first4=Audrey|last5=Tshibanda|first5=Luaba|last6=Thonnard|first6=Marie|last7=Charland-Verville|first7=Vanessa|last8=Kirsch|first8=Murielle|last9=Laureys|first9=Steven|last10=Demertzi|first10=Athena|title=Resting state networks and consciousness. Alterations of multiple resting state network connectivity in physiological, pharmacological and pathological consciousness states.|journal=Frontiers in Psychology|date=2012|volume=3|pages=295|doi=10.3389/fpsyg.2012.00295|pmid=22969735|pmc=3427917|doi-access=free}}</ref><ref name="Yeo">{{cite journal|last1=Yeo|first1=B. T. Thomas|last2=Krienen|first2=Fenna M.|last3=Sepulcre|first3=Jorge|last4=Sabuncu|first4=Mert R.|last5=Lashkari|first5=Danial|last6=Hollinshead|first6=Marisa|last7=Roffman|first7=Joshua L.|last8=Smoller|first8=Jordan W.|last9=Zöllei|first9=Lilla|last10=Polimeni|first10=Jonathan R.|last11=Fischl|first11=Bruce|last12=Liu|first12=Hesheng|last13=Buckner|first13=Randy L.|title=The organization of the human cerebral cortex estimated by intrinsic functional connectivity|journal=Journal of Neurophysiology|date=2011-09-01|volume=106|issue=3|pages=1125–1165|doi=10.1152/jn.00338.2011|pmid=21653723|pmc=3174820|bibcode=2011NatSD...2E0031H }}</ref><ref name="Shafiei">{{cite journal|last1=Shafiei|first1=Golia|last2=Zeighami|first2=Yashar|last3=Clark|first3=Crystal A.|last4=Coull|first4=Jennifer T.|last5=Nagano-Saito|first5=Atsuko|last6=Leyton|first6=Marco|last7=Dagher|first7=Alain|last8=Mišić|first8=Bratislav|title=Dopamine Signaling Modulates the Stability and Integration of Intrinsic Brain Networks|journal=Cerebral Cortex|date=2018-10-01|volume=29|issue=1|pages=397–409|doi=10.1093/cercor/bhy264|pmid=30357316|pmc=6294404 }}</ref><ref name="Bailey">{{cite journal|last1=Bailey|first1=Stephen K.|last2=Aboud|first2=Katherine S.|last3=Nguyen|first3=Tin Q.|last4=Cutting|first4=Laurie E.|title=Applying a network framework to the neurobiology of reading and dyslexia|journal=Journal of Neurodevelopmental Disorders|date=13 December 2018|volume=10|issue=1|page=37|doi=10.1186/s11689-018-9251-z|pmid=30541433|pmc=6291929 }}</ref><br />
<br />
===默认模式网络（内侧额顶骨）===<br />
*默认模式网络在个体清醒和休息时都处于活跃状态。当个体专注于内部导向任务，比如做白日梦、展望未来、提取回忆和'''<font color="#ff8000">心智理论Theory of mind</font>'''任务时，默认模式网络会被优先激活。它与专注于外部视觉信号的大脑系统成负相关。对默认模式网络的研究目前是所谓网络之中最为广泛的。<ref name="Bressler" /><ref name="Bassett" /><ref name=":10" /><ref name="Riedl" /><ref name="Yuan" /><ref name="Bell" /><ref name="Heine" /><ref name="Yeo" /><ref name="Shafiei" /><ref name="Bailey" /><br />
<br />
===Salience (Midcingulo-Insular)===<br />
{{Main|Salience network}}<br />
*The salience network consists of several structures, including the anterior (bilateral) insula, dorsal anterior cingulate cortex, and three subcortical structures which are the ventral striatum, substantia nigra/ventral tegmental region.<ref name=":11">{{Cite journal|last1=Steimke|first1=Rosa|last2=Nomi|first2=Jason S.|last3=Calhoun|first3=Vince D.|last4=Stelzel|first4=Christine|last5=Paschke|first5=Lena M.|last6=Gaschler|first6=Robert|last7=Goschke|first7=Thomas|last8=Walter|first8=Henrik|last9=Uddin|first9=Lucina Q.|date=2017-12-01|title=Salience network dynamics underlying successful resistance of temptation|journal=Social Cognitive and Affective Neuroscience|language=en|volume=12|issue=12|pages=1928–1939|doi=10.1093/scan/nsx123|pmid=29048582|pmc=5716209|issn=1749-5016|doi-access=free}}</ref><ref name=":0">{{Citation|last=Menon|first=V.|title=Brain Mapping|chapter=Salience Network|date=2015-01-01|chapter-url=http://www.sciencedirect.com/science/article/pii/B978012397025100052X|pages=597–611|editor-last=Toga|editor-first=Arthur W.|publisher=Academic Press|isbn=978-0-12-397316-0|access-date=2019-12-08|doi=10.1016/B978-0-12-397025-1.00052-X}}</ref> It plays the key role of monitoring the [[Salience (neuroscience)|salience]] of external inputs and internal brain events.<ref name="Riedl" /><ref name="Bressler" /><ref name="Bassett" /><ref name="Yuan" /><ref name="Heine" /><ref name="Yeo" /><ref name="Shafiei" /> Specifically, it aids in directing attention by identifying important biological and cognitive events.<ref name=":0" /><ref name="Bailey" /><br />
*This network includes the ventral attention network, which primarily includes the [[temporoparietal junction]] and the ventral [[frontal cortex]] of the right hemisphere.<ref name="Uddin2019" /><ref name="Vossel" /> These areas respond when behaviorally relevant stimuli occur unexpectedly.<ref name="Vossel" /> The ventral attention network is inhibited during focused attention in which top-down processing is being used, such as when visually searching for something. This response may prevent goal-driven attention from being distracted by non-relevant stimuli. It becomes active again when the target or relevant information about the target is found.<ref name="Vossel" /><ref name=":12">{{Cite journal|last1=Shulman|first1=Gordon L.|last2=McAvoy|first2=Mark P.|last3=Cowan|first3=Melanie C.|last4=Astafiev|first4=Serguei V.|last5=Tansy|first5=Aaron P.|last6=d'Avossa|first6=Giovanni|last7=Corbetta|first7=Maurizio|date=2003-11-01|title=Quantitative Analysis of Attention and Detection Signals During Visual Search|journal=Journal of Neurophysiology|volume=90|issue=5|pages=3384–3397|doi=10.1152/jn.00343.2003|pmid=12917383|issn=0022-3077}}</ref><br />
<br />
<br />
===突显网络（扣带回-岛叶）===<br />
*突显网络由'''<font color="#32CD32"> 前(双)岛、前扣带回皮层和三个皮层下结构（腹侧纹状体、黑质/腹侧被盖区）</font>'''组成，<ref name=":11" /><ref name=":0" />它在监测外部输入和内部脑活动的'''<font color="#ff8000">凸显程度Salience</font>'''中发挥着关键作用。<ref name="Riedl" /><ref name="Bressler" /><ref name="Bassett" /><ref name="Yuan" /><ref name="Heine" /><ref name="Yeo" /><ref name="Shafiei" />具体来说，突显网络通过识别重要的生理、认知活动来帮助引导注意力。<ref name=":0" /><ref name="Bailey" /><br />
*突显网络包括主要由右半球的'''<font color="#ff8000">颞顶联合区Temporoparietal junction</font>'''和腹侧'''<font color="#ff8000">额叶Frontal cortex</font>'''皮层组成的腹侧注意网络。<ref name="Uddin2019" /><ref name="Vossel" />当行为相关的刺激意外发生时，这些区域会对此做出反应。在自上而下加工注意焦点的过程中（例如使用视觉搜索某件物品时），腹侧注意网络会受到抑制。这种抑制反应可以防止目标驱动的注意力被不相关的刺激分散。当找到目标或相关信息时，突显网络会被再次激活。<ref name="Vossel" /><ref name=":12" /><br />
<br />
===Attention (Dorsal frontoparietal)===<br />
{{Main|Dorsal attention network}}<br />
*This network is involved in the voluntary, top-down deployment of attention.<ref name="Riedl" /><ref name="Yuan" /><ref name="Bell" /><ref name="Yeo" /><ref name="Shafiei" /><ref name="Vossel">{{cite journal|last1=Vossel|first1=Simone|last2=Geng|first2=Joy J.|last3=Fink|first3=Gereon R.|title=Dorsal and Ventral Attention Systems: Distinct Neural Circuits but Collaborative Roles|journal=The Neuroscientist|date=2014|volume=20|issue=2|pages=150–159|doi=10.1177/1073858413494269|pmid=23835449|pmc=4107817}}</ref><ref name="Hutton">{{cite journal|last1=Hutton|first1=John S.|last2=Dudley|first2=Jonathan|last3=Horowitz-Kraus|first3=Tzipi|last4=DeWitt|first4=Tom|last5=Holland|first5=Scott K.|title=Functional Connectivity of Attention, Visual, and Language Networks During Audio, Illustrated, and Animated Stories in Preschool-Age Children|journal=Brain Connectivity|date=1 September 2019|volume=9|issue=7|pages=580–592|doi=10.1089/brain.2019.0679|pmid=31144523|pmc=6775495|ref=Hutton}}</ref> Within the dorsal attention network, the intraparietal sulcus and frontal eye fields influence the visual areas of the brain. These influencing factors allow for the orientation of attention.<ref name=":13">{{Cite journal|last1=Fox|first1=Michael D.|last2=Corbetta|first2=Maurizio|last3=Snyder|first3=Abraham Z.|last4=Vincent|first4=Justin L.|last5=Raichle|first5=Marcus E.|date=2006-06-27|title=Spontaneous neuronal activity distinguishes human dorsal and ventral attention systems|journal=Proceedings of the National Academy of Sciences|language=en|volume=103|issue=26|pages=10046–10051|doi=10.1073/pnas.0604187103|issn=0027-8424|pmid=16788060|pmc=1480402|bibcode=2006PNAS..10310046F|doi-access=free}}</ref><ref name="Vossel" /><ref name="Bailey" /><br />
<br />
===注意网络（背侧额顶骨）===<br />
*这个网络参与了自发的、自上而下的注意力分配。<ref name="Riedl" /><ref name="Yuan" /><ref name="Bell" /><ref name="Yeo" /><ref name="Shafiei" /><ref name="Vossel" /><ref name="Hutton" />在背侧注意网络中，顶内沟和额眼影响着大脑的视觉区域。这些影响因素决定了注意力的方向。<ref name=":13" /><ref name="Vossel" /><ref name="Bailey" /><br />
<br />
=== Control (Lateral frontoparietal)===<br />
{{Main|Frontoparietal network}}<br />
*This network initiates and modulates cognitive control and comprises 18 sub-regions of the brain.<ref name=":14">{{Cite journal|last1=Scolari|first1=Miranda|last2=Seidl-Rathkopf|first2=Katharina N|last3=Kastner|first3=Sabine|date=2015-02-01|title=Functions of the human frontoparietal attention network: Evidence from neuroimaging|journal=Current Opinion in Behavioral Sciences|series=Cognitive control|volume=1|pages=32–39|doi=10.1016/j.cobeha.2014.08.003|issn=2352-1546|pmid=27398396|pmc=4936532}}</ref> There is a strong correlation between fluid intelligence and the involvement of the fronto-parietal network with other networks.<ref name=":15">{{Cite journal|last1=Marek|first1=Scott|last2=Dosenbach|first2=Nico U. F.|date=June 2018|title=The frontoparietal network: function, electrophysiology, and importance of individual precision mapping|journal=Dialogues in Clinical Neuroscience|volume=20|issue=2|pages=133–140|doi=10.31887/DCNS.2018.20.2/smarek|issn=1294-8322|pmc=6136121|pmid=30250390}}</ref><br />
<br />
*Versions of this network have also been called the central executive (or executive control) network and the cognitive control network.<ref name="Uddin2019" /><br />
<br />
===控制网络（侧额顶骨）===<br />
*这个网络参与了认知控制的启动与调节，它包括了大脑的18个亚区。<ref name=":14" />额顶网络与其他网络的参与程度和流体智力之间存在着有很强的相关性。<ref name=":15" /><br />
*在其它版本中，这种网络也被称为中央执行(或执行控制)网络和认知控制网络。<ref name="Uddin2019" /><br />
<br />
===Sensorimotor or Somatomotor (Pericentral)===<br />
{{Main|Sensorimotor network}}<br />
*This network processes somatosensory information and coordinates motion.<ref name="Heine" /><ref name="Yeo" /><ref name="Shafiei" /><ref name="Bassett" /><ref name="Yuan" /> The [[auditory cortex]] may be included.<ref name="Uddin2019" /><ref name="Yeo" /><br />
<br />
===感觉运动网络（中央区域） ===<br />
*这个网络参与了躯体感觉信息的加工和运动的协调。<ref name="Heine" /><ref name="Yeo" /><ref name="Shafiei" /><ref name="Bassett" /><ref name="Yuan" />'''<font color="#ff8000">听觉皮层Auditory cortex</font>'''可能也包括在内。<ref name="Uddin2019" /><ref name="Yeo" /><br />
<br />
===Visual (Occipital)===<br />
{{See|Visual cortex}}<br />
*This network handles visual information processing.<ref name="Yang">{{cite journal|last1=Yang|first1=Yan-li|last2=Deng|first2=Hong-xia|last3=Xing|first3=Gui-yang|last4=Xia|first4=Xiao-luan|last5=Li|first5=Hai-fang|title=Brain functional network connectivity based on a visual task: visual information processing-related brain regions are significantly activated in the task state|journal=Neural Regeneration Research|date=2015|volume=10|issue=2|pages=298–307|doi=10.4103/1673-5374.152386|pmid=25883631|pmc=4392680 }}</ref><br />
<br />
===视觉网络（枕部）===<br />
*这个网络处理视觉信息。<ref name="Yang" /><br />
<br />
==Other networks==<br />
Different methods and data have identified several other brain networks, many of which greatly overlap or are subsets of more well-characterized core networks.<ref name="Uddin2019" /><br />
*Limbic<ref name="Bassett" /><ref name="Yeo" /><ref name="Bailey" /><br />
*Auditory<ref name="Yuan" /><ref name="Heine" /><br />
*Right/left executive<ref name="Yuan" /><ref name="Heine" /><br />
*Cerebellar<ref name="Bell" /><ref name="Heine" /><br />
* Spatial attention<ref name="Riedl" /><ref name="Bressler" /><br />
*Language<ref name="Bressler" /><ref name="Hutton" /><br />
*Lateral visual<ref name="Yuan" /><ref name="Bell" /><ref name="Heine" /><br />
*Temporal<ref name="Yeo" /><ref name="Shafiei" /><br />
*Visual perception/imagery<ref name="Hutton" /><br />
<br />
==其他网络==<br />
不同的方法和数据已经能够确定上述核心网络以外的几个大脑网络，其中许多网络之间存在极大的重叠，部分网络实际上是核心网络更具特点的子集。<ref name="Uddin2019" /><br />
*边缘网络<ref name="Bassett" /><ref name="Yeo" /><ref name="Bailey" /><br />
*听觉网络<ref name="Yuan" /><ref name="Heine" /><br />
*右/左执行网络<ref name="Yuan" /><ref name="Heine" /><br />
*小脑网络<ref name="Bell" /><ref name="Heine" /><br />
*空间注意网络<ref name="Riedl" /><ref name="Bressler" /><br />
*语言网络<ref name="Bressler" /><ref name="Hutton" /><br />
*外侧视觉网络<ref name="Yuan" /><ref name="Bell" /><ref name="Heine" /><br />
*颞骨网络<ref name="Yeo" /><ref name="Shafiei" /><br />
*视知觉/图像网络<ref name="Hutton" /><br />
<br />
==See also==<br />
*[[Complex network]]<br />
*[[Neural network]]<br />
<br />
==相关词条==<br />
*[[复杂网络]]<br />
*[[神经网络]]<br />
<br />
==References==<br />
{{reflist|30em}}<br />
<br />
{{Human connectomics}}<br />
<br />
[[Category:Neuroscience]]<br />
[[Category:Neural coding]]<br />
[[Category:Neural circuits]]<br />
[[Category:Neurophysiology]]<br />
<br />
Category:Neuroscience<br />
Category:Neural coding<br />
Category:Neural circuits<br />
Category:Neurophysiology<br />
<br />
类别: 神经科学类别: 神经编码类别: 神经回路类别: 神经生理学<br />
<br />
<noinclude><br />
<br />
<small>This page was moved from [[wikipedia:en:Large-scale brain network]]. Its edit history can be viewed at [[大规模脑网络/edithistory]]</small></noinclude><br />
<br />
[[Category:待整理页面]]</div>Shenky20https://wiki.swarma.org/index.php?title=%E5%A4%A7%E8%A7%84%E6%A8%A1%E8%84%91%E7%BD%91%E7%BB%9C&diff=29994大规模脑网络2022-04-04T09:52:06Z<p>Shenky20：</p>
<hr />
<div>此词条由神经动力学模型读书会词条梳理志愿者Shenky20翻译审校，翻译字数共800，未经专家审核，带来阅读不便，请见谅。<br />
<br />
{{Short description|Collections of brain regions}}<br />
'''Large-scale brain networks''' (also known as '''intrinsic brain networks''') are collections of widespread [[brain regions]] showing [[Resting state fMRI#Functional|functional connectivity]] by statistical analysis of the [[Functional magnetic resonance imaging|fMRI]] [[BOLD signal]]<ref name=Riedl>{{cite journal|last1=Riedl|first1=Valentin|last2=Utz|first2=Lukas|last3=Castrillón|first3=Gabriel|last4=Grimmer|first4=Timo|last5=Rauschecker|first5=Josef P.|last6=Ploner|first6=Markus|last7=Friston|first7=Karl J.|last8=Drzezga|first8=Alexander|last9=Sorg|first9=Christian|title=Metabolic connectivity mapping reveals effective connectivity in the resting human brain|journal=PNAS|date=January 12, 2016|volume=113|issue=2|pages=428–433|doi=10.1073/pnas.1513752113|pmid=26712010|pmc=4720331|bibcode=2016PNAS..113..428R|doi-access=free}}</ref> or other recording methods such as [[Electroencephalography|EEG]],<ref name=":1">{{Cite journal|last1=Foster|first1=Brett L.|last2=Parvizi|first2=Josef|date=2012-03-01|title=Resting oscillations and cross-frequency coupling in the human posteromedial cortex|journal=NeuroImage|volume=60|issue=1|pages=384–391|doi=10.1016/j.neuroimage.2011.12.019|pmid=22227048|issn=1053-8119|pmc=3596417}}</ref> [[Positron emission tomography|PET]]<ref name=":2">{{Cite journal|last1=Buckner|first1=Randy L.|last2=Andrews‐Hanna|first2=Jessica R.|last3=Schacter|first3=Daniel L.|date=2008|title=The Brain's Default Network|journal=Annals of the New York Academy of Sciences|language=en|volume=1124|issue=1|pages=1–38|doi=10.1196/annals.1440.011|pmid=18400922|issn=1749-6632|bibcode=2008NYASA1124....1B|s2cid=3167595}}</ref> and [[Magnetoencephalography|MEG]].<ref name=":3">{{Cite journal|last1=Morris|first1=Peter G.|last2=Smith|first2=Stephen M.|last3=Barnes|first3=Gareth R.|last4=Stephenson|first4=Mary C.|last5=Hale|first5=Joanne R.|last6=Price|first6=Darren|last7=Luckhoo|first7=Henry|last8=Woolrich|first8=Mark|last9=Brookes|first9=Matthew J.|date=2011-10-04|title=Investigating the electrophysiological basis of resting state networks using magnetoencephalography|journal=Proceedings of the National Academy of Sciences|language=en|volume=108|issue=40|pages=16783–16788|doi=10.1073/pnas.1112685108|issn=0027-8424|pmid=21930901|pmc=3189080|bibcode=2011PNAS..10816783B|doi-access=free}}</ref> An emerging paradigm in neuroscience is that cognitive tasks are performed not by individual brain regions working in isolation but by networks consisting of several discrete brain regions that are said to be "functionally connected". Functional connectivity networks may be found using algorithms such as [[cluster analysis]], spatial [[independent component analysis]] (ICA), seed based, and others.<ref name="Petersen">{{cite journal|last1=Petersen|first1=Steven|last2=Sporns|first2=Olaf|title=Brain Networks and Cognitive Architectures|journal=Neuron|date=October 2015|volume=88|issue=1|pages=207–219|doi=10.1016/j.neuron.2015.09.027|pmid=26447582|pmc=4598639 }}</ref> Synchronized brain regions may also be identified using long-range synchronization of the EEG, MEG, or other dynamic brain signals.<ref name=Bressler>{{cite journal|last1=Bressler|first1=Steven L.|last2=Menon|first2=Vinod|s2cid=5967761|title=Large scale brain networks in cognition: emerging methods and principles|journal=Trends in Cognitive Sciences|date=June 2010|volume=14|issue=6|pages=233–290|doi=10.1016/j.tics.2010.04.004|url=http://www.cell.com/trends/cognitive-sciences/issue?pii=S1364-6613(10)X0005-5|accessdate=24 January 2016|pmid=20493761}}</ref><br />
<br />
'''<font color="#ff8000">大规模脑网络Large-scale brain networks</font>'''(也称为'''<font color="#ff8000">内在大脑网络Intrinsic brain networks</font>''')是在对基于'''<font color="#ff8000">血氧水平依赖效应BOLD</font>'''的'''<font color="#ff8000">功能性磁共振成像fMRI</font>'''信号<ref name="Riedl" />的统计分析或其他记录方法(如'''<font color="#ff8000">脑电图EEG<ref name=":1" /></font>'''、'''<font color="#ff8000">正电子发射断层扫描技术PET<ref name=":2" /></font>'''和'''<font color="#ff8000">脑磁图MEG<ref name=":3" /></font>''')中，表现出'''<font color="#ff8000">功能连接Functional connectivity</font>'''的'''<font color="#ff8000">脑区Brain regions</font>'''的集合。根据神经科学中一个新出现的范式，认知任务不是由单个脑区独立执行的，而是由几个互不相连的脑区“功能连接”组成的网络执行的。功能连接网络可以通过'''<font color="#ff8000">数据聚类Cluster analysis</font>'''、空间'''<font color="#ff8000">独立元素分析ICA</font>'''、种子点方法等算法来发现。<ref name="Petersen" />同步的脑区也可以用脑电图、脑磁图或其他动态脑信号的远程同步来识别。<ref name="Bressler" /><br />
<br />
The set of identified brain areas that are linked together in a large-scale network varies with cognitive function.<ref name="Bressler2">{{cite journal|last1=Bressler|first1=Steven L.|title=Neurocognitive networks|journal=Scholarpedia|volume=3|issue=2|pages=1567|doi=10.4249/scholarpedia.1567|year=2008|bibcode=2008SchpJ...3.1567B|doi-access=free}}</ref> When the cognitive state is not explicit (i.e., the subject is at "rest"), the large-scale brain network is a [[Resting state fMRI|resting state]] network (RSN). As a physical system with graph-like properties,<ref name="Bressler" /> a large-scale brain network has both nodes and edges and cannot be identified simply by the co-activation of brain areas. In recent decades, the analysis of brain networks was made feasible by advances in imaging techniques as well as new tools from [[graph theory]] and [[Dynamical systems theory|dynamical systems]].<br />
<br />
<br />
大规模脑网络中,连接在一起的脑区集合因认知功能的不同而不同。<ref name="Bressler2" />当认知状态不明确(即主体处于“静止”状态)时，大规模脑网络是一个'''<font color="#ff8000">静息状态Resting state</font>'''网络(RSN)。作为一个具有图形特征的物理系统，<ref name="Bressler" />大规模脑网络既有节点又有边，不能简单地通过脑区的共同激活来识别。近几十年来，成像技术不断进步，同时'''<font color="#ff8000">图论Graph theory</font>'''、'''<font color="#ff8000">动力学系统Dynamical systems</font>'''领域出现了新的技术手段，这使得脑网络分析变得可行。<br />
<br />
Large-scale brain networks are identified by their function and provide a coherent framework for understanding [[cognition]] by offering a neural model of how different cognitive functions emerge when different sets of brain regions join together as self-organized coalitions. The number and composition of the coalitions will vary with the algorithm and parameters used to identify them.<ref name="Yeo" /><ref name=":4">{{cite journal|last1=Abou Elseoud|first1=Ahmed|last2=Littow|first2=Harri|last3=Remes|first3=Jukka|last4=Starck|first4=Tuomo|last5=Nikkinen|first5=Juha|last6=Nissilä|first6=Juuso|last7=Timonen|first7=Markku|last8=Tervonen|first8=Osmo|last9=Kiviniemi1|first9=Vesa|title=Group-ICA Model Order Highlights Patterns of Functional Brain Connectivity|journal= Frontiers in Systems Neuroscience|date=2011-06-03|volume=5|pages=37|doi=10.3389/fnsys.2011.00037|pmid=21687724|pmc=3109774 |doi-access=free}}</ref> In one model, there is only the [[default mode network]] and the [[task-positive network]], but most current analyses show several networks, from a small handful to 17.<ref name="Yeo" /> The most common and stable networks are enumerated below. The regions participating in a functional network may be dynamically reconfigured.<ref name="Petersen" /><ref name="Bassett">{{cite journal|last1=Bassett|first1=Daniella|last2=Bertolero|first2=Max|title=How Matter Becomes Mind|journal=Scientific American|date=July 2019|volume=321|issue=1|page=32|url=https://www.scientificamerican.com/|accessdate=23 June 2019 }}</ref><br />
<br />
<br />
大规模脑网络是通过其功能来进行识别的。通过研究大规模脑网络建立神经模型，对不同脑区组合所形成的自组织联合体如何实现不同的'''<font color="#ff8000">认知Cognition</font>'''功能进行解释，就能够为认知理解提供一个连贯的框架。识别算法和参数的不同会导致所识别出的上述联合体的数量和组成有所不同。<ref name="Yeo" /><ref name=":4" />一个模型理论认为，符合上述条件的神经模型只包含'''<font color="#ff8000">默认模式网络Default mode network</font>'''和'''<font color="#ff8000">任务激活网络Task-positive network</font>'''，但目前'''<font color="#32CD32">大多数分析理论都包括从几个到17个不等的网络。<ref name="Yeo" /></font>'''下面列举了最常见且稳定的网络。'''<font color="#32CD32"> 人脑</font>'''可以动态地重新配置参与功能网络的脑区。<ref name="Petersen" /><ref name="Bassett" /> <br />
<br />
Disruptions in activity in various networks have been implicated in neuropsychiatric disorders such as [[Depression (mood)|depression]], [[Alzheimer's disease|Alzheimer's]], [[Autism-spectrum disorder|autism spectrum disorder]], [[schizophrenia]], [[ADHD]]<ref name=":5">{{cite journal |last1=Griffiths |first1=Kristi R. |last2=Braund |first2=Taylor A. |last3=Kohn |first3=Michael R. |last4=Clarke |first4=Simon |last5=Williams |first5=Leanne M. |last6=Korgaonkar |first6=Mayuresh S. |title=Structural brain network topology underpinning ADHD and response to methylphenidate treatment |journal=Translational Psychiatry |date=2 March 2021 |volume=11 |issue=1 |pages=1–9 |doi=10.1038/s41398-021-01278-x | pmc=7925571 |pmid=33654073 |url=https://www.nature.com/articles/s41398-021-01278-x#citeas |access-date=16 November 2021}}</ref> and [[bipolar disorder]].<ref name=":6">{{Cite journal|url=https://www.researchgate.net/publication/51639686|title=Large-scale brain networks and psychopathology: A unifying triple network model|last=Menon|first=Vinod|s2cid=26653572|journal=Trends in Cognitive Sciences|date=2011-09-09|volume=15|issue=10|pages=483–506|doi=10.1016/j.tics.2011.08.003|pmid=21908230}}</ref><br />
<br />
<br />
脑网络活动的中断与诸多神经精神疾病密切相关，如'''<font color="#ff8000">抑郁症Depression</font>'''、'''<font color="#ff8000">老年痴呆症Alzheimer's</font>'''、'''<font color="#ff8000">自闭症谱系障碍Autism spectrum disorder</font>'''、'''<font color="#ff8000">精神分裂症Schizophrenia</font>'''、'''<font color="#ff8000">多动症ADHD<ref name=":5" /></font>'''和'''<font color="#ff8000">躁郁症Bipolar disorder<ref name=":6" /></font>'''。<br />
<br />
==Core networks==<br />
[[File:Heine2012x3010.png|thumb|An example that identified 10 large-scale brain networks from [[resting state fMRI]] activity through [[independent component analysis]].<ref name="Heine" />|链接=Special:FilePath/Heine2012x3010.png]]<br />
Because brain networks can be identified at various different resolutions and with various different neurobiological properties, there is no such thing as a universal atlas of brain networks that fits all circumstances.<ref name=":7">{{cite journal|last1=Eickhoff|first1=SB|last2=Yeo|first2=BTT|last3=Genon|first3=S|title=Imaging-based parcellations of the human brain.|journal=Nature Reviews. Neuroscience|date=November 2018|volume=19|issue=11|pages=672–686|doi=10.1038/s41583-018-0071-7|pmid=30305712|s2cid=52954265|url=http://juser.fz-juelich.de/record/856633/files/Eickhoff_Yeo_Genon_NRN_MainManuscriptInclFigures.pdf}}</ref> While acknowledging this problem, Uddin, Yeo, and Spreng proposed in 2019<ref name="Uddin2019">{{cite journal|last1=Uddin|first1=LQ|last2=Yeo|first2=BTT|last3=Spreng|first3=RN|title=Towards a Universal Taxonomy of Macro-scale Functional Human Brain Networks.|journal=Brain Topography|date=November 2019|volume=32|issue=6|pages=926–942|doi=10.1007/s10548-019-00744-6|pmid=31707621|pmc=7325607}}</ref> that the following six networks should be defined as core networks based on converging evidences from multiple studies<ref name=":8">{{cite journal|last1=Doucet|first1=GE|last2=Lee|first2=WH|last3=Frangou|first3=S|title=Evaluation of the spatial variability in the major resting-state networks across human brain functional atlases.|journal=Human Brain Mapping|date=2019-10-15|volume=40|issue=15|pages=4577–4587|doi=10.1002/hbm.24722|pmid=31322303|pmc=6771873}}</ref><ref name="Yeo" /><ref name=":9">{{cite journal|last1=Smith|first1=SM|last2=Fox|first2=PT|last3=Miller|first3=KL|last4=Glahn|first4=DC|last5=Fox|first5=PM|last6=Mackay|first6=CE|last7=Filippini|first7=N|last8=Watkins|first8=KE|last9=Toro|first9=R|last10=Laird|first10=AR|last11=Beckmann|first11=CF|title=Correspondence of the brain's functional architecture during activation and rest.|journal=Proceedings of the National Academy of Sciences of the United States of America|date=2009-08-04|volume=106|issue=31|pages=13040–5|doi=10.1073/pnas.0905267106|pmid=19620724|pmc=2722273|bibcode=2009PNAS..10613040S|doi-access=free}}</ref> to facilitate communication between researchers.<br />
<br />
<br />
==核心网络==<br />
【图1：An example that identified 10 large-scale brain networks from resting state fMRI activity through independent component analysis.<ref name="Heine" />这是一个通过独立元素分析方法从静息状态脑功能性磁共振成像信号中分辨出10个大规模脑网络的例子。<ref name="Heine" />】<br />
<br />
<br />
因为脑网络可以在不同分辨率条件下、使用不同的神经生物学特性来识别，所以没有能够适用于所有情况的通用脑网络图谱。<ref name=":7" />在承认这个问题的同时，Uddin，Yeo，和Spreng在2019年提出<ref name="Uddin2019" />，综合考虑来自多个研究的证据<ref name=":8" /><ref name="Yeo" /><ref name=":9" />，以下六个网络应该被定义为核心网络，以促进研究人员之间的交流。<br />
<br />
===Default Mode (Medial frontoparietal)===<br />
{{Main|Default mode network}}<br />
*The default mode network is active when an individual is awake and at rest. It preferentially activates when individuals focus on internally-oriented tasks such as daydreaming, envisioning the future, retrieving memories, and [[theory of mind]]. It is negatively correlated with brain systems that focus on external visual signals. It is the most widely researched network.<ref name="Bressler" /><ref name="Bassett" /><ref name=":10">{{Cite journal|date=2012-08-15|title=The serendipitous discovery of the brain's default network|journal=NeuroImage|language=en|volume=62|issue=2|pages=1137–1145|doi=10.1016/j.neuroimage.2011.10.035|pmid=22037421|issn=1053-8119|last1=Buckner|first1=Randy L.|s2cid=9880586}}</ref><ref name="Riedl" /><ref name="Yuan"><br />
{{cite journal|last1=Yuan|first1=Rui|last2=Di|first2=Xin|last3=Taylor|first3=Paul A.|last4=Gohel|first4=Suril|last5=Tsai|first5=Yuan-Hsiung|last6=Biswal|first6=Bharat B.|title=Functional topography of the thalamocortical system in human|journal=Brain Structure and Function|date=30 April 2015|doi=10.1007/s00429-015-1018-7|pmid=25924563|pmc=6363530|volume=221|issue=4|pages=1971–1984}}</ref><ref name="Bell">{{cite journal|last1=Bell|first1=Peter T.|last2=Shine|first2=James M.|title=Estimating Large-Scale Network Convergence in the Human Functional Connectome|journal=Brain Connectivity|date=2015-11-09|volume=5|issue=9|doi=10.1089/brain.2015.0348|pmid=26005099|pages=565–74}}</ref><ref name="Heine">{{cite journal|last1=Heine|first1=Lizette|last2=Soddu|first2=Andrea|last3=Gomez|first3=Francisco|last4=Vanhaudenhuyse|first4=Audrey|last5=Tshibanda|first5=Luaba|last6=Thonnard|first6=Marie|last7=Charland-Verville|first7=Vanessa|last8=Kirsch|first8=Murielle|last9=Laureys|first9=Steven|last10=Demertzi|first10=Athena|title=Resting state networks and consciousness. Alterations of multiple resting state network connectivity in physiological, pharmacological and pathological consciousness states.|journal=Frontiers in Psychology|date=2012|volume=3|pages=295|doi=10.3389/fpsyg.2012.00295|pmid=22969735|pmc=3427917|doi-access=free}}</ref><ref name="Yeo">{{cite journal|last1=Yeo|first1=B. T. Thomas|last2=Krienen|first2=Fenna M.|last3=Sepulcre|first3=Jorge|last4=Sabuncu|first4=Mert R.|last5=Lashkari|first5=Danial|last6=Hollinshead|first6=Marisa|last7=Roffman|first7=Joshua L.|last8=Smoller|first8=Jordan W.|last9=Zöllei|first9=Lilla|last10=Polimeni|first10=Jonathan R.|last11=Fischl|first11=Bruce|last12=Liu|first12=Hesheng|last13=Buckner|first13=Randy L.|title=The organization of the human cerebral cortex estimated by intrinsic functional connectivity|journal=Journal of Neurophysiology|date=2011-09-01|volume=106|issue=3|pages=1125–1165|doi=10.1152/jn.00338.2011|pmid=21653723|pmc=3174820|bibcode=2011NatSD...2E0031H }}</ref><ref name="Shafiei">{{cite journal|last1=Shafiei|first1=Golia|last2=Zeighami|first2=Yashar|last3=Clark|first3=Crystal A.|last4=Coull|first4=Jennifer T.|last5=Nagano-Saito|first5=Atsuko|last6=Leyton|first6=Marco|last7=Dagher|first7=Alain|last8=Mišić|first8=Bratislav|title=Dopamine Signaling Modulates the Stability and Integration of Intrinsic Brain Networks|journal=Cerebral Cortex|date=2018-10-01|volume=29|issue=1|pages=397–409|doi=10.1093/cercor/bhy264|pmid=30357316|pmc=6294404 }}</ref><ref name="Bailey">{{cite journal|last1=Bailey|first1=Stephen K.|last2=Aboud|first2=Katherine S.|last3=Nguyen|first3=Tin Q.|last4=Cutting|first4=Laurie E.|title=Applying a network framework to the neurobiology of reading and dyslexia|journal=Journal of Neurodevelopmental Disorders|date=13 December 2018|volume=10|issue=1|page=37|doi=10.1186/s11689-018-9251-z|pmid=30541433|pmc=6291929 }}</ref><br />
<br />
<br />
===默认模式网络（内侧额顶骨）===<br />
*默认模式网络在个体清醒和休息时都处于活跃状态。当个体专注于内部导向任务，比如做白日梦、展望未来、提取回忆和'''<font color="#ff8000">心智理论Theory of mind</font>'''任务时，默认模式网络会被优先激活。它与专注于外部视觉信号的大脑系统成负相关。对默认模式网络的研究目前是所谓网络之中最为广泛的。<ref name="Bressler" /><ref name="Bassett" /><ref name=":10" /><ref name="Riedl" /><ref name="Yuan" /><ref name="Bell" /><ref name="Heine" /><ref name="Yeo" /><ref name="Shafiei" /><ref name="Bailey" /><br />
<br />
===Salience (Midcingulo-Insular)===<br />
{{Main|Salience network}}<br />
*The salience network consists of several structures, including the anterior (bilateral) insula, dorsal anterior cingulate cortex, and three subcortical structures which are the ventral striatum, substantia nigra/ventral tegmental region.<ref name=":11">{{Cite journal|last1=Steimke|first1=Rosa|last2=Nomi|first2=Jason S.|last3=Calhoun|first3=Vince D.|last4=Stelzel|first4=Christine|last5=Paschke|first5=Lena M.|last6=Gaschler|first6=Robert|last7=Goschke|first7=Thomas|last8=Walter|first8=Henrik|last9=Uddin|first9=Lucina Q.|date=2017-12-01|title=Salience network dynamics underlying successful resistance of temptation|journal=Social Cognitive and Affective Neuroscience|language=en|volume=12|issue=12|pages=1928–1939|doi=10.1093/scan/nsx123|pmid=29048582|pmc=5716209|issn=1749-5016|doi-access=free}}</ref><ref name=":0">{{Citation|last=Menon|first=V.|title=Brain Mapping|chapter=Salience Network|date=2015-01-01|chapter-url=http://www.sciencedirect.com/science/article/pii/B978012397025100052X|pages=597–611|editor-last=Toga|editor-first=Arthur W.|publisher=Academic Press|isbn=978-0-12-397316-0|access-date=2019-12-08|doi=10.1016/B978-0-12-397025-1.00052-X}}</ref> It plays the key role of monitoring the [[Salience (neuroscience)|salience]] of external inputs and internal brain events.<ref name="Riedl" /><ref name="Bressler" /><ref name="Bassett" /><ref name="Yuan" /><ref name="Heine" /><ref name="Yeo" /><ref name="Shafiei" /> Specifically, it aids in directing attention by identifying important biological and cognitive events.<ref name=":0" /><ref name="Bailey" /><br />
*This network includes the ventral attention network, which primarily includes the [[temporoparietal junction]] and the ventral [[frontal cortex]] of the right hemisphere.<ref name="Uddin2019" /><ref name="Vossel" /> These areas respond when behaviorally relevant stimuli occur unexpectedly.<ref name="Vossel" /> The ventral attention network is inhibited during focused attention in which top-down processing is being used, such as when visually searching for something. This response may prevent goal-driven attention from being distracted by non-relevant stimuli. It becomes active again when the target or relevant information about the target is found.<ref name="Vossel" /><ref name=":12">{{Cite journal|last1=Shulman|first1=Gordon L.|last2=McAvoy|first2=Mark P.|last3=Cowan|first3=Melanie C.|last4=Astafiev|first4=Serguei V.|last5=Tansy|first5=Aaron P.|last6=d'Avossa|first6=Giovanni|last7=Corbetta|first7=Maurizio|date=2003-11-01|title=Quantitative Analysis of Attention and Detection Signals During Visual Search|journal=Journal of Neurophysiology|volume=90|issue=5|pages=3384–3397|doi=10.1152/jn.00343.2003|pmid=12917383|issn=0022-3077}}</ref><br />
<br />
<br />
===突显网络（扣带回-岛叶）===<br />
*突显网络由'''<font color="#32CD32"> 前(双)岛、前扣带回皮层和三个皮层下结构（腹侧纹状体、黑质/腹侧被盖区）</font>'''组成，<ref name=":11" /><ref name=":0" />它在监测外部输入和内部脑活动的'''<font color="#ff8000">凸显程度Salience</font>'''中发挥着关键作用。<ref name="Riedl" /><ref name="Bressler" /><ref name="Bassett" /><ref name="Yuan" /><ref name="Heine" /><ref name="Yeo" /><ref name="Shafiei" />具体来说，突显网络通过识别重要的生理、认知活动来帮助引导注意力。<ref name=":0" /><ref name="Bailey" /><br />
*突显网络包括主要由右半球的'''<font color="#ff8000">颞顶联合区Temporoparietal junction</font>'''和腹侧'''<font color="#ff8000">额叶Frontal cortex</font>'''皮层组成的腹侧注意网络。<ref name="Uddin2019" /><ref name="Vossel" />当行为相关的刺激意外发生时，这些区域会对此做出反应。在自上而下加工注意焦点的过程中（例如使用视觉搜索某件物品时），腹侧注意网络会受到抑制。这种抑制反应可以防止目标驱动的注意力被不相关的刺激分散。当找到目标或相关信息时，突显网络会被再次激活。<ref name="Vossel" /><ref name=":12" /><br />
<br />
===Attention (Dorsal frontoparietal)===<br />
{{Main|Dorsal attention network}}<br />
*This network is involved in the voluntary, top-down deployment of attention.<ref name="Riedl" /><ref name="Yuan" /><ref name="Bell" /><ref name="Yeo" /><ref name="Shafiei" /><ref name="Vossel">{{cite journal|last1=Vossel|first1=Simone|last2=Geng|first2=Joy J.|last3=Fink|first3=Gereon R.|title=Dorsal and Ventral Attention Systems: Distinct Neural Circuits but Collaborative Roles|journal=The Neuroscientist|date=2014|volume=20|issue=2|pages=150–159|doi=10.1177/1073858413494269|pmid=23835449|pmc=4107817}}</ref><ref name="Hutton">{{cite journal|last1=Hutton|first1=John S.|last2=Dudley|first2=Jonathan|last3=Horowitz-Kraus|first3=Tzipi|last4=DeWitt|first4=Tom|last5=Holland|first5=Scott K.|title=Functional Connectivity of Attention, Visual, and Language Networks During Audio, Illustrated, and Animated Stories in Preschool-Age Children|journal=Brain Connectivity|date=1 September 2019|volume=9|issue=7|pages=580–592|doi=10.1089/brain.2019.0679|pmid=31144523|pmc=6775495|ref=Hutton}}</ref> Within the dorsal attention network, the intraparietal sulcus and frontal eye fields influence the visual areas of the brain. These influencing factors allow for the orientation of attention.<ref name=":13">{{Cite journal|last1=Fox|first1=Michael D.|last2=Corbetta|first2=Maurizio|last3=Snyder|first3=Abraham Z.|last4=Vincent|first4=Justin L.|last5=Raichle|first5=Marcus E.|date=2006-06-27|title=Spontaneous neuronal activity distinguishes human dorsal and ventral attention systems|journal=Proceedings of the National Academy of Sciences|language=en|volume=103|issue=26|pages=10046–10051|doi=10.1073/pnas.0604187103|issn=0027-8424|pmid=16788060|pmc=1480402|bibcode=2006PNAS..10310046F|doi-access=free}}</ref><ref name="Vossel" /><ref name="Bailey" /><br />
<br />
<br />
===注意网络（背侧额顶骨）===<br />
*这个网络参与了自发的、自上而下的注意力分配。<ref name="Riedl" /><ref name="Yuan" /><ref name="Bell" /><ref name="Yeo" /><ref name="Shafiei" /><ref name="Vossel" /><ref name="Hutton" />在背侧注意网络中，顶内沟和额眼影响着大脑的视觉区域。这些影响因素决定了注意力的方向。<ref name=":13" /><ref name="Vossel" /><ref name="Bailey" /><br />
<br />
=== Control (Lateral frontoparietal)===<br />
{{Main|Frontoparietal network}}<br />
*This network initiates and modulates cognitive control and comprises 18 sub-regions of the brain.<ref name=":14">{{Cite journal|last1=Scolari|first1=Miranda|last2=Seidl-Rathkopf|first2=Katharina N|last3=Kastner|first3=Sabine|date=2015-02-01|title=Functions of the human frontoparietal attention network: Evidence from neuroimaging|journal=Current Opinion in Behavioral Sciences|series=Cognitive control|volume=1|pages=32–39|doi=10.1016/j.cobeha.2014.08.003|issn=2352-1546|pmid=27398396|pmc=4936532}}</ref> There is a strong correlation between fluid intelligence and the involvement of the fronto-parietal network with other networks.<ref name=":15">{{Cite journal|last1=Marek|first1=Scott|last2=Dosenbach|first2=Nico U. F.|date=June 2018|title=The frontoparietal network: function, electrophysiology, and importance of individual precision mapping|journal=Dialogues in Clinical Neuroscience|volume=20|issue=2|pages=133–140|doi=10.31887/DCNS.2018.20.2/smarek|issn=1294-8322|pmc=6136121|pmid=30250390}}</ref><br />
<br />
*Versions of this network have also been called the central executive (or executive control) network and the cognitive control network.<ref name="Uddin2019" /><br />
<br />
<br />
===控制网络（侧额顶骨）===<br />
*这个网络参与了认知控制的启动与调节，它包括了大脑的18个亚区。<ref name=":14" />额顶网络与其他网络的参与程度和流体智力之间存在着有很强的相关性。<ref name=":15" /><br />
*在其它版本中，这种网络也被称为中央执行(或执行控制)网络和认知控制网络。<ref name="Uddin2019" /><br />
<br />
===Sensorimotor or Somatomotor (Pericentral)===<br />
{{Main|Sensorimotor network}}<br />
*This network processes somatosensory information and coordinates motion.<ref name="Heine" /><ref name="Yeo" /><ref name="Shafiei" /><ref name="Bassett" /><ref name="Yuan" /> The [[auditory cortex]] may be included.<ref name="Uddin2019" /><ref name="Yeo" /><br />
<br />
===感觉运动网络（中央区域） ===<br />
*这个网络参与了躯体感觉信息的加工和运动的协调。<ref name="Heine" /><ref name="Yeo" /><ref name="Shafiei" /><ref name="Bassett" /><ref name="Yuan" />'''<font color="#ff8000">听觉皮层Auditory cortex</font>'''可能也包括在内。<ref name="Uddin2019" /><ref name="Yeo" /><br />
<br />
===Visual (Occipital)===<br />
{{See|Visual cortex}}<br />
*This network handles visual information processing.<ref name="Yang">{{cite journal|last1=Yang|first1=Yan-li|last2=Deng|first2=Hong-xia|last3=Xing|first3=Gui-yang|last4=Xia|first4=Xiao-luan|last5=Li|first5=Hai-fang|title=Brain functional network connectivity based on a visual task: visual information processing-related brain regions are significantly activated in the task state|journal=Neural Regeneration Research|date=2015|volume=10|issue=2|pages=298–307|doi=10.4103/1673-5374.152386|pmid=25883631|pmc=4392680 }}</ref><br />
<br />
===视觉网络（枕部）===<br />
*这个网络处理视觉信息。<ref name="Yang" /><br />
<br />
==Other networks==<br />
Different methods and data have identified several other brain networks, many of which greatly overlap or are subsets of more well-characterized core networks.<ref name="Uddin2019" /><br />
*Limbic<ref name="Bassett" /><ref name="Yeo" /><ref name="Bailey" /><br />
*Auditory<ref name="Yuan" /><ref name="Heine" /><br />
*Right/left executive<ref name="Yuan" /><ref name="Heine" /><br />
*Cerebellar<ref name="Bell" /><ref name="Heine" /><br />
* Spatial attention<ref name="Riedl" /><ref name="Bressler" /><br />
*Language<ref name="Bressler" /><ref name="Hutton" /><br />
*Lateral visual<ref name="Yuan" /><ref name="Bell" /><ref name="Heine" /><br />
*Temporal<ref name="Yeo" /><ref name="Shafiei" /><br />
*Visual perception/imagery<ref name="Hutton" /><br />
<br />
==其他网络==<br />
不同的方法和数据已经能够确定上述核心网络以外的几个大脑网络，其中许多网络之间存在极大的重叠，部分网络实际上是核心网络更具特点的子集。<ref name="Uddin2019" /><br />
*边缘网络<ref name="Bassett" /><ref name="Yeo" /><ref name="Bailey" /><br />
*听觉网络<ref name="Yuan" /><ref name="Heine" /><br />
*右/左执行网络<ref name="Yuan" /><ref name="Heine" /><br />
*小脑网络<ref name="Bell" /><ref name="Heine" /><br />
*空间注意网络<ref name="Riedl" /><ref name="Bressler" /><br />
*语言网络<ref name="Bressler" /><ref name="Hutton" /><br />
*外侧视觉网络<ref name="Yuan" /><ref name="Bell" /><ref name="Heine" /><br />
*颞骨网络<ref name="Yeo" /><ref name="Shafiei" /><br />
*颞视知觉/图像网络<ref name="Hutton" /><br />
<br />
==See also==<br />
*[[Complex network]]<br />
*[[Neural network]]<br />
<br />
==相关词条==<br />
*[[复杂网络]]<br />
*[[神经网络]]<br />
<br />
==References==<br />
{{reflist|30em}}<br />
<br />
{{Human connectomics}}<br />
<br />
[[Category:Neuroscience]]<br />
[[Category:Neural coding]]<br />
[[Category:Neural circuits]]<br />
[[Category:Neurophysiology]]<br />
<br />
Category:Neuroscience<br />
Category:Neural coding<br />
Category:Neural circuits<br />
Category:Neurophysiology<br />
<br />
类别: 神经科学类别: 神经编码类别: 神经回路类别: 神经生理学<br />
<br />
<noinclude><br />
<br />
<small>This page was moved from [[wikipedia:en:Large-scale brain network]]. Its edit history can be viewed at [[大规模脑网络/edithistory]]</small></noinclude><br />
<br />
[[Category:待整理页面]]</div>Shenky20https://wiki.swarma.org/index.php?title=%E5%A4%A7%E8%A7%84%E6%A8%A1%E8%84%91%E7%BD%91%E7%BB%9C&diff=29993大规模脑网络2022-04-04T09:47:25Z<p>Shenky20：</p>
<hr />
<div>此词条由神经动力学模型读书会词条梳理志愿者Shenky20翻译审校，翻译字数共800，未经专家审核，带来阅读不便，请见谅。<br />
<br />
{{Short description|Collections of brain regions}}<br />
'''Large-scale brain networks''' (also known as '''intrinsic brain networks''') are collections of widespread [[brain regions]] showing [[Resting state fMRI#Functional|functional connectivity]] by statistical analysis of the [[Functional magnetic resonance imaging|fMRI]] [[BOLD signal]]<ref name=Riedl>{{cite journal|last1=Riedl|first1=Valentin|last2=Utz|first2=Lukas|last3=Castrillón|first3=Gabriel|last4=Grimmer|first4=Timo|last5=Rauschecker|first5=Josef P.|last6=Ploner|first6=Markus|last7=Friston|first7=Karl J.|last8=Drzezga|first8=Alexander|last9=Sorg|first9=Christian|title=Metabolic connectivity mapping reveals effective connectivity in the resting human brain|journal=PNAS|date=January 12, 2016|volume=113|issue=2|pages=428–433|doi=10.1073/pnas.1513752113|pmid=26712010|pmc=4720331|bibcode=2016PNAS..113..428R|doi-access=free}}</ref> or other recording methods such as [[Electroencephalography|EEG]],<ref name=":1">{{Cite journal|last1=Foster|first1=Brett L.|last2=Parvizi|first2=Josef|date=2012-03-01|title=Resting oscillations and cross-frequency coupling in the human posteromedial cortex|journal=NeuroImage|volume=60|issue=1|pages=384–391|doi=10.1016/j.neuroimage.2011.12.019|pmid=22227048|issn=1053-8119|pmc=3596417}}</ref> [[Positron emission tomography|PET]]<ref name=":2">{{Cite journal|last1=Buckner|first1=Randy L.|last2=Andrews‐Hanna|first2=Jessica R.|last3=Schacter|first3=Daniel L.|date=2008|title=The Brain's Default Network|journal=Annals of the New York Academy of Sciences|language=en|volume=1124|issue=1|pages=1–38|doi=10.1196/annals.1440.011|pmid=18400922|issn=1749-6632|bibcode=2008NYASA1124....1B|s2cid=3167595}}</ref> and [[Magnetoencephalography|MEG]].<ref name=":3">{{Cite journal|last1=Morris|first1=Peter G.|last2=Smith|first2=Stephen M.|last3=Barnes|first3=Gareth R.|last4=Stephenson|first4=Mary C.|last5=Hale|first5=Joanne R.|last6=Price|first6=Darren|last7=Luckhoo|first7=Henry|last8=Woolrich|first8=Mark|last9=Brookes|first9=Matthew J.|date=2011-10-04|title=Investigating the electrophysiological basis of resting state networks using magnetoencephalography|journal=Proceedings of the National Academy of Sciences|language=en|volume=108|issue=40|pages=16783–16788|doi=10.1073/pnas.1112685108|issn=0027-8424|pmid=21930901|pmc=3189080|bibcode=2011PNAS..10816783B|doi-access=free}}</ref> An emerging paradigm in neuroscience is that cognitive tasks are performed not by individual brain regions working in isolation but by networks consisting of several discrete brain regions that are said to be "functionally connected". Functional connectivity networks may be found using algorithms such as [[cluster analysis]], spatial [[independent component analysis]] (ICA), seed based, and others.<ref name="Petersen">{{cite journal|last1=Petersen|first1=Steven|last2=Sporns|first2=Olaf|title=Brain Networks and Cognitive Architectures|journal=Neuron|date=October 2015|volume=88|issue=1|pages=207–219|doi=10.1016/j.neuron.2015.09.027|pmid=26447582|pmc=4598639 }}</ref> Synchronized brain regions may also be identified using long-range synchronization of the EEG, MEG, or other dynamic brain signals.<ref name=Bressler>{{cite journal|last1=Bressler|first1=Steven L.|last2=Menon|first2=Vinod|s2cid=5967761|title=Large scale brain networks in cognition: emerging methods and principles|journal=Trends in Cognitive Sciences|date=June 2010|volume=14|issue=6|pages=233–290|doi=10.1016/j.tics.2010.04.004|url=http://www.cell.com/trends/cognitive-sciences/issue?pii=S1364-6613(10)X0005-5|accessdate=24 January 2016|pmid=20493761}}</ref><br />
<br />
<br />
<br />
'''<font color="#ff8000">大规模脑网络Large-scale brain networks</font>'''(也称为'''<font color="#ff8000">内在大脑网络Intrinsic brain networks</font>''')是在对基于'''<font color="#ff8000">血氧水平依赖效应BOLD</font>'''的'''<font color="#ff8000">功能性磁共振成像fMRI</font>'''信号<ref name="Riedl" />的统计分析或其他记录方法(如'''<font color="#ff8000">脑电图EEG<ref name=":1" /></font>'''、'''<font color="#ff8000">正电子发射断层扫描技术PET<ref name=":2" /></font>'''和'''<font color="#ff8000">脑磁图MEG<ref name=":3" /></font>''')中，表现出'''<font color="#ff8000">功能连接Functional connectivity</font>'''的'''<font color="#ff8000">脑区Brain regions</font>'''的集合。根据神经科学中一个新出现的范式，认知任务不是由单个脑区独立执行的，而是由几个互不相连的脑区“功能连接”组成的网络执行的。功能连接网络可以通过'''<font color="#ff8000">数据聚类Cluster analysis</font>'''、空间'''<font color="#ff8000">独立元素分析ICA</font>'''、种子点方法等算法来发现。<ref name="Petersen" />同步的脑区也可以用脑电图、脑磁图或其他动态脑信号的远程同步来识别。<ref name="Bressler" /><br />
<br />
The set of identified brain areas that are linked together in a large-scale network varies with cognitive function.<ref name="Bressler2">{{cite journal|last1=Bressler|first1=Steven L.|title=Neurocognitive networks|journal=Scholarpedia|volume=3|issue=2|pages=1567|doi=10.4249/scholarpedia.1567|year=2008|bibcode=2008SchpJ...3.1567B|doi-access=free}}</ref> When the cognitive state is not explicit (i.e., the subject is at "rest"), the large-scale brain network is a [[Resting state fMRI|resting state]] network (RSN). As a physical system with graph-like properties,<ref name="Bressler" /> a large-scale brain network has both nodes and edges and cannot be identified simply by the co-activation of brain areas. In recent decades, the analysis of brain networks was made feasible by advances in imaging techniques as well as new tools from [[graph theory]] and [[Dynamical systems theory|dynamical systems]].<br />
<br />
<br />
大规模脑网络中,连接在一起的脑区集合因认知功能的不同而不同。<ref name="Bressler2" />当认知状态不明确(即主体处于“静止”状态)时，大规模脑网络是一个'''<font color="#ff8000">静息状态Resting state</font>'''网络(RSN)。作为一个具有图形特征的物理系统，<ref name="Bressler" />大规模脑网络既有节点又有边，不能简单地通过脑区的共同激活来识别。近几十年来，成像技术不断进步，同时'''<font color="#ff8000">图论Graph theory</font>'''、'''<font color="#ff8000">动力学系统Dynamical systems</font>'''领域出现了新的技术手段，这使得脑网络分析变得可行。<br />
<br />
Large-scale brain networks are identified by their function and provide a coherent framework for understanding [[cognition]] by offering a neural model of how different cognitive functions emerge when different sets of brain regions join together as self-organized coalitions. The number and composition of the coalitions will vary with the algorithm and parameters used to identify them.<ref name="Yeo" /><ref name=":4">{{cite journal|last1=Abou Elseoud|first1=Ahmed|last2=Littow|first2=Harri|last3=Remes|first3=Jukka|last4=Starck|first4=Tuomo|last5=Nikkinen|first5=Juha|last6=Nissilä|first6=Juuso|last7=Timonen|first7=Markku|last8=Tervonen|first8=Osmo|last9=Kiviniemi1|first9=Vesa|title=Group-ICA Model Order Highlights Patterns of Functional Brain Connectivity|journal= Frontiers in Systems Neuroscience|date=2011-06-03|volume=5|pages=37|doi=10.3389/fnsys.2011.00037|pmid=21687724|pmc=3109774 |doi-access=free}}</ref> In one model, there is only the [[default mode network]] and the [[task-positive network]], but most current analyses show several networks, from a small handful to 17.<ref name="Yeo" /> The most common and stable networks are enumerated below. The regions participating in a functional network may be dynamically reconfigured.<ref name="Petersen" /><ref name="Bassett">{{cite journal|last1=Bassett|first1=Daniella|last2=Bertolero|first2=Max|title=How Matter Becomes Mind|journal=Scientific American|date=July 2019|volume=321|issue=1|page=32|url=https://www.scientificamerican.com/|accessdate=23 June 2019 }}</ref><br />
<br />
<br />
大规模脑网络是通过其功能来进行识别的。通过研究大规模脑网络建立神经模型，对不同脑区组合所形成的自组织联合体如何实现不同的'''<font color="#ff8000">认知Cognition</font>'''功能进行解释，就能够为认知理解提供一个连贯的框架。识别算法和参数的不同会导致所识别出的上述联合体的数量和组成有所不同。<ref name="Yeo" /><ref name=":4" />一个模型理论认为，符合上述条件的神经模型只包含'''<font color="#ff8000">默认模式网络Default mode network</font>'''和'''<font color="#ff8000">任务激活网络Task-positive network</font>'''，但目前'''<font color="#32CD32">大多数分析理论都包括从几个到17个不等的网络。<ref name="Yeo" /></font>'''下面列举了最常见且稳定的网络。'''<font color="#32CD32"> 人脑</font>'''可以动态地重新配置参与功能网络的脑区。<ref name="Petersen" /><ref name="Bassett" /> <br />
<br />
Disruptions in activity in various networks have been implicated in neuropsychiatric disorders such as [[Depression (mood)|depression]], [[Alzheimer's disease|Alzheimer's]], [[Autism-spectrum disorder|autism spectrum disorder]], [[schizophrenia]], [[ADHD]]<ref name=":5">{{cite journal |last1=Griffiths |first1=Kristi R. |last2=Braund |first2=Taylor A. |last3=Kohn |first3=Michael R. |last4=Clarke |first4=Simon |last5=Williams |first5=Leanne M. |last6=Korgaonkar |first6=Mayuresh S. |title=Structural brain network topology underpinning ADHD and response to methylphenidate treatment |journal=Translational Psychiatry |date=2 March 2021 |volume=11 |issue=1 |pages=1–9 |doi=10.1038/s41398-021-01278-x | pmc=7925571 |pmid=33654073 |url=https://www.nature.com/articles/s41398-021-01278-x#citeas |access-date=16 November 2021}}</ref> and [[bipolar disorder]].<ref name=":6">{{Cite journal|url=https://www.researchgate.net/publication/51639686|title=Large-scale brain networks and psychopathology: A unifying triple network model|last=Menon|first=Vinod|s2cid=26653572|journal=Trends in Cognitive Sciences|date=2011-09-09|volume=15|issue=10|pages=483–506|doi=10.1016/j.tics.2011.08.003|pmid=21908230}}</ref><br />
<br />
<br />
脑网络活动的中断与诸多神经精神疾病密切相关，如'''<font color="#ff8000">抑郁症Depression</font>'''、'''<font color="#ff8000">老年痴呆症Alzheimer's</font>'''、'''<font color="#ff8000">自闭症谱系障碍Autism spectrum disorder</font>'''、'''<font color="#ff8000">精神分裂症Schizophrenia</font>'''、'''<font color="#ff8000">多动症ADHD<ref name=":5" /></font>'''和'''<font color="#ff8000">躁郁症Bipolar disorder<ref name=":6" /></font>'''。<br />
<br />
==Core networks==<br />
[[File:Heine2012x3010.png|thumb|An example that identified 10 large-scale brain networks from [[resting state fMRI]] activity through [[independent component analysis]].<ref name="Heine" />|链接=Special:FilePath/Heine2012x3010.png]]<br />
Because brain networks can be identified at various different resolutions and with various different neurobiological properties, there is no such thing as a universal atlas of brain networks that fits all circumstances.<ref name=":7">{{cite journal|last1=Eickhoff|first1=SB|last2=Yeo|first2=BTT|last3=Genon|first3=S|title=Imaging-based parcellations of the human brain.|journal=Nature Reviews. Neuroscience|date=November 2018|volume=19|issue=11|pages=672–686|doi=10.1038/s41583-018-0071-7|pmid=30305712|s2cid=52954265|url=http://juser.fz-juelich.de/record/856633/files/Eickhoff_Yeo_Genon_NRN_MainManuscriptInclFigures.pdf}}</ref> While acknowledging this problem, Uddin, Yeo, and Spreng proposed in 2019<ref name="Uddin2019">{{cite journal|last1=Uddin|first1=LQ|last2=Yeo|first2=BTT|last3=Spreng|first3=RN|title=Towards a Universal Taxonomy of Macro-scale Functional Human Brain Networks.|journal=Brain Topography|date=November 2019|volume=32|issue=6|pages=926–942|doi=10.1007/s10548-019-00744-6|pmid=31707621|pmc=7325607}}</ref> that the following six networks should be defined as core networks based on converging evidences from multiple studies<ref name=":8">{{cite journal|last1=Doucet|first1=GE|last2=Lee|first2=WH|last3=Frangou|first3=S|title=Evaluation of the spatial variability in the major resting-state networks across human brain functional atlases.|journal=Human Brain Mapping|date=2019-10-15|volume=40|issue=15|pages=4577–4587|doi=10.1002/hbm.24722|pmid=31322303|pmc=6771873}}</ref><ref name="Yeo" /><ref name=":9">{{cite journal|last1=Smith|first1=SM|last2=Fox|first2=PT|last3=Miller|first3=KL|last4=Glahn|first4=DC|last5=Fox|first5=PM|last6=Mackay|first6=CE|last7=Filippini|first7=N|last8=Watkins|first8=KE|last9=Toro|first9=R|last10=Laird|first10=AR|last11=Beckmann|first11=CF|title=Correspondence of the brain's functional architecture during activation and rest.|journal=Proceedings of the National Academy of Sciences of the United States of America|date=2009-08-04|volume=106|issue=31|pages=13040–5|doi=10.1073/pnas.0905267106|pmid=19620724|pmc=2722273|bibcode=2009PNAS..10613040S|doi-access=free}}</ref> to facilitate communication between researchers.<br />
<br />
<br />
==核心网络==<br />
=<nowiki>= = 核心网络 = =</nowiki>= <br />
【图1：An example that identified 10 large-scale brain networks from resting state fMRI activity through independent component analysis.<ref name="Heine" />这是一个通过独立元素分析方法从静息状态脑功能性磁共振成像信号中分辨出10个大规模脑网络的例子。<ref name="Heine" />】<br />
<br />
<br />
因为脑网络可以在不同分辨率条件下、使用不同的神经生物学特性来识别，所以没有能够适用于所有情况的通用脑网络图谱。<ref name=":7" />在承认这个问题的同时，Uddin，Yeo，和Spreng在2019年提出<ref name="Uddin2019" />，综合考虑来自多个研究的证据<ref name=":8" /><ref name="Yeo" /><ref name=":9" />，以下六个网络应该被定义为核心网络，以促进研究人员之间的交流。<br />
<br />
===Default Mode (Medial frontoparietal)===<br />
{{Main|Default mode network}}<br />
*The default mode network is active when an individual is awake and at rest. It preferentially activates when individuals focus on internally-oriented tasks such as daydreaming, envisioning the future, retrieving memories, and [[theory of mind]]. It is negatively correlated with brain systems that focus on external visual signals. It is the most widely researched network.<ref name="Bressler" /><ref name="Bassett" /><ref name=":10">{{Cite journal|date=2012-08-15|title=The serendipitous discovery of the brain's default network|journal=NeuroImage|language=en|volume=62|issue=2|pages=1137–1145|doi=10.1016/j.neuroimage.2011.10.035|pmid=22037421|issn=1053-8119|last1=Buckner|first1=Randy L.|s2cid=9880586}}</ref><ref name="Riedl" /><ref name="Yuan"><br />
{{cite journal|last1=Yuan|first1=Rui|last2=Di|first2=Xin|last3=Taylor|first3=Paul A.|last4=Gohel|first4=Suril|last5=Tsai|first5=Yuan-Hsiung|last6=Biswal|first6=Bharat B.|title=Functional topography of the thalamocortical system in human|journal=Brain Structure and Function|date=30 April 2015|doi=10.1007/s00429-015-1018-7|pmid=25924563|pmc=6363530|volume=221|issue=4|pages=1971–1984}}</ref><ref name="Bell">{{cite journal|last1=Bell|first1=Peter T.|last2=Shine|first2=James M.|title=Estimating Large-Scale Network Convergence in the Human Functional Connectome|journal=Brain Connectivity|date=2015-11-09|volume=5|issue=9|doi=10.1089/brain.2015.0348|pmid=26005099|pages=565–74}}</ref><ref name="Heine">{{cite journal|last1=Heine|first1=Lizette|last2=Soddu|first2=Andrea|last3=Gomez|first3=Francisco|last4=Vanhaudenhuyse|first4=Audrey|last5=Tshibanda|first5=Luaba|last6=Thonnard|first6=Marie|last7=Charland-Verville|first7=Vanessa|last8=Kirsch|first8=Murielle|last9=Laureys|first9=Steven|last10=Demertzi|first10=Athena|title=Resting state networks and consciousness. Alterations of multiple resting state network connectivity in physiological, pharmacological and pathological consciousness states.|journal=Frontiers in Psychology|date=2012|volume=3|pages=295|doi=10.3389/fpsyg.2012.00295|pmid=22969735|pmc=3427917|doi-access=free}}</ref><ref name="Yeo">{{cite journal|last1=Yeo|first1=B. T. Thomas|last2=Krienen|first2=Fenna M.|last3=Sepulcre|first3=Jorge|last4=Sabuncu|first4=Mert R.|last5=Lashkari|first5=Danial|last6=Hollinshead|first6=Marisa|last7=Roffman|first7=Joshua L.|last8=Smoller|first8=Jordan W.|last9=Zöllei|first9=Lilla|last10=Polimeni|first10=Jonathan R.|last11=Fischl|first11=Bruce|last12=Liu|first12=Hesheng|last13=Buckner|first13=Randy L.|title=The organization of the human cerebral cortex estimated by intrinsic functional connectivity|journal=Journal of Neurophysiology|date=2011-09-01|volume=106|issue=3|pages=1125–1165|doi=10.1152/jn.00338.2011|pmid=21653723|pmc=3174820|bibcode=2011NatSD...2E0031H }}</ref><ref name="Shafiei">{{cite journal|last1=Shafiei|first1=Golia|last2=Zeighami|first2=Yashar|last3=Clark|first3=Crystal A.|last4=Coull|first4=Jennifer T.|last5=Nagano-Saito|first5=Atsuko|last6=Leyton|first6=Marco|last7=Dagher|first7=Alain|last8=Mišić|first8=Bratislav|title=Dopamine Signaling Modulates the Stability and Integration of Intrinsic Brain Networks|journal=Cerebral Cortex|date=2018-10-01|volume=29|issue=1|pages=397–409|doi=10.1093/cercor/bhy264|pmid=30357316|pmc=6294404 }}</ref><ref name="Bailey">{{cite journal|last1=Bailey|first1=Stephen K.|last2=Aboud|first2=Katherine S.|last3=Nguyen|first3=Tin Q.|last4=Cutting|first4=Laurie E.|title=Applying a network framework to the neurobiology of reading and dyslexia|journal=Journal of Neurodevelopmental Disorders|date=13 December 2018|volume=10|issue=1|page=37|doi=10.1186/s11689-018-9251-z|pmid=30541433|pmc=6291929 }}</ref><br />
<br />
<br />
===默认模式网络（内侧额顶骨）===<br />
=<nowiki>= = 默认模式网络（内侧额顶叶） = =</nowiki>= <br />
*默认模式网络在个体清醒和休息时都处于活跃状态。当个体专注于内部导向任务，比如做白日梦、展望未来、提取回忆和'''<font color="#ff8000">心智理论Theory of mind</font>'''任务时，默认模式网络会被优先激活。它与专注于外部视觉信号的大脑系统成负相关。对默认模式网络的研究目前是所谓网络之中最为广泛的。<ref name="Bressler" /><ref name="Bassett" /><ref name=":10" /><ref name="Riedl" /><ref name="Yuan" /><ref name="Bell" /><ref name="Heine" /><ref name="Yeo" /><ref name="Shafiei" /><ref name="Bailey" /><br />
<br />
=== Salience (Midcingulo-Insular)===<br />
{{Main|Salience network}}<br />
*The salience network consists of several structures, including the anterior (bilateral) insula, dorsal anterior cingulate cortex, and three subcortical structures which are the ventral striatum, substantia nigra/ventral tegmental region.<ref name=":11">{{Cite journal|last1=Steimke|first1=Rosa|last2=Nomi|first2=Jason S.|last3=Calhoun|first3=Vince D.|last4=Stelzel|first4=Christine|last5=Paschke|first5=Lena M.|last6=Gaschler|first6=Robert|last7=Goschke|first7=Thomas|last8=Walter|first8=Henrik|last9=Uddin|first9=Lucina Q.|date=2017-12-01|title=Salience network dynamics underlying successful resistance of temptation|journal=Social Cognitive and Affective Neuroscience|language=en|volume=12|issue=12|pages=1928–1939|doi=10.1093/scan/nsx123|pmid=29048582|pmc=5716209|issn=1749-5016|doi-access=free}}</ref><ref name=":0">{{Citation|last=Menon|first=V.|title=Brain Mapping|chapter=Salience Network|date=2015-01-01|chapter-url=http://www.sciencedirect.com/science/article/pii/B978012397025100052X|pages=597–611|editor-last=Toga|editor-first=Arthur W.|publisher=Academic Press|isbn=978-0-12-397316-0|access-date=2019-12-08|doi=10.1016/B978-0-12-397025-1.00052-X}}</ref> It plays the key role of monitoring the [[Salience (neuroscience)|salience]] of external inputs and internal brain events.<ref name="Riedl" /><ref name="Bressler" /><ref name="Bassett" /><ref name="Yuan" /><ref name="Heine" /><ref name="Yeo" /><ref name="Shafiei" /> Specifically, it aids in directing attention by identifying important biological and cognitive events.<ref name=":0" /><ref name="Bailey" /><br />
*This network includes the ventral attention network, which primarily includes the [[temporoparietal junction]] and the ventral [[frontal cortex]] of the right hemisphere.<ref name="Uddin2019" /><ref name="Vossel" /> These areas respond when behaviorally relevant stimuli occur unexpectedly.<ref name="Vossel" /> The ventral attention network is inhibited during focused attention in which top-down processing is being used, such as when visually searching for something. This response may prevent goal-driven attention from being distracted by non-relevant stimuli. It becomes active again when the target or relevant information about the target is found.<ref name="Vossel" /><ref name=":12">{{Cite journal|last1=Shulman|first1=Gordon L.|last2=McAvoy|first2=Mark P.|last3=Cowan|first3=Melanie C.|last4=Astafiev|first4=Serguei V.|last5=Tansy|first5=Aaron P.|last6=d'Avossa|first6=Giovanni|last7=Corbetta|first7=Maurizio|date=2003-11-01|title=Quantitative Analysis of Attention and Detection Signals During Visual Search|journal=Journal of Neurophysiology|volume=90|issue=5|pages=3384–3397|doi=10.1152/jn.00343.2003|pmid=12917383|issn=0022-3077}}</ref><br />
<br />
<br />
===突显网络（扣带回-岛叶）===<br />
=<nowiki>= = 突显网络（扣带回-岛叶）= =</nowiki>= <br />
*突显网络由'''<font color="#32CD32"> 前(双)岛、前扣带回皮层和三个皮层下结构（腹侧纹状体、黑质/腹侧被盖区）</font>'''组成，<ref name=":11" /><ref name=":0" />它在监测外部输入和内部脑活动的'''<font color="#ff8000">凸显程度Salience</font>'''中发挥着关键作用。<ref name="Riedl" /><ref name="Bressler" /><ref name="Bassett" /><ref name="Yuan" /><ref name="Heine" /><ref name="Yeo" /><ref name="Shafiei" />具体来说，突显网络通过识别重要的生理、认知活动来帮助引导注意力。<ref name=":0" /><ref name="Bailey" /><br />
*突显网络包括主要由右半球的'''<font color="#ff8000">颞顶联合区Temporoparietal junction</font>'''和腹侧'''<font color="#ff8000">额叶Frontal cortex</font>'''皮层组成的腹侧注意网络。<ref name="Uddin2019" /><ref name="Vossel" />当行为相关的刺激意外发生时，这些区域会对此做出反应。在自上而下加工注意焦点的过程中（例如使用视觉搜索某件物品时），腹侧注意网络会受到抑制。这种抑制反应可以防止目标驱动的注意力被不相关的刺激分散。当找到目标或相关信息时，突显网络会被再次激活。<ref name="Vossel" /><ref name=":12" /><br />
<br />
===Attention (Dorsal frontoparietal)===<br />
{{Main|Dorsal attention network}}<br />
*This network is involved in the voluntary, top-down deployment of attention.<ref name="Riedl" /><ref name="Yuan" /><ref name="Bell" /><ref name="Yeo" /><ref name="Shafiei" /><ref name="Vossel">{{cite journal|last1=Vossel|first1=Simone|last2=Geng|first2=Joy J.|last3=Fink|first3=Gereon R.|title=Dorsal and Ventral Attention Systems: Distinct Neural Circuits but Collaborative Roles|journal=The Neuroscientist|date=2014|volume=20|issue=2|pages=150–159|doi=10.1177/1073858413494269|pmid=23835449|pmc=4107817}}</ref><ref name="Hutton">{{cite journal|last1=Hutton|first1=John S.|last2=Dudley|first2=Jonathan|last3=Horowitz-Kraus|first3=Tzipi|last4=DeWitt|first4=Tom|last5=Holland|first5=Scott K.|title=Functional Connectivity of Attention, Visual, and Language Networks During Audio, Illustrated, and Animated Stories in Preschool-Age Children|journal=Brain Connectivity|date=1 September 2019|volume=9|issue=7|pages=580–592|doi=10.1089/brain.2019.0679|pmid=31144523|pmc=6775495|ref=Hutton}}</ref> Within the dorsal attention network, the intraparietal sulcus and frontal eye fields influence the visual areas of the brain. These influencing factors allow for the orientation of attention.<ref name=":13">{{Cite journal|last1=Fox|first1=Michael D.|last2=Corbetta|first2=Maurizio|last3=Snyder|first3=Abraham Z.|last4=Vincent|first4=Justin L.|last5=Raichle|first5=Marcus E.|date=2006-06-27|title=Spontaneous neuronal activity distinguishes human dorsal and ventral attention systems|journal=Proceedings of the National Academy of Sciences|language=en|volume=103|issue=26|pages=10046–10051|doi=10.1073/pnas.0604187103|issn=0027-8424|pmid=16788060|pmc=1480402|bibcode=2006PNAS..10310046F|doi-access=free}}</ref><ref name="Vossel" /><ref name="Bailey" /><br />
<br />
<br />
===注意网络（背侧额顶骨）===<br />
=<nowiki>= = 注意网络(背侧额顶骨) = =</nowiki>= <br />
*这个网络参与了自发的、自上而下的注意力分配。<ref name="Riedl" /><ref name="Yuan" /><ref name="Bell" /><ref name="Yeo" /><ref name="Shafiei" /><ref name="Vossel" /><ref name="Hutton" />在背侧注意网络中，顶内沟和额眼影响着大脑的视觉区域。这些影响因素决定了注意力的方向。<ref name=":13" /><ref name="Vossel" /><ref name="Bailey" /><br />
<br />
===Control (Lateral frontoparietal)===<br />
{{Main|Frontoparietal network}}<br />
*This network initiates and modulates cognitive control and comprises 18 sub-regions of the brain.<ref name=":14">{{Cite journal|last1=Scolari|first1=Miranda|last2=Seidl-Rathkopf|first2=Katharina N|last3=Kastner|first3=Sabine|date=2015-02-01|title=Functions of the human frontoparietal attention network: Evidence from neuroimaging|journal=Current Opinion in Behavioral Sciences|series=Cognitive control|volume=1|pages=32–39|doi=10.1016/j.cobeha.2014.08.003|issn=2352-1546|pmid=27398396|pmc=4936532}}</ref> There is a strong correlation between fluid intelligence and the involvement of the fronto-parietal network with other networks.<ref name=":15">{{Cite journal|last1=Marek|first1=Scott|last2=Dosenbach|first2=Nico U. F.|date=June 2018|title=The frontoparietal network: function, electrophysiology, and importance of individual precision mapping|journal=Dialogues in Clinical Neuroscience|volume=20|issue=2|pages=133–140|doi=10.31887/DCNS.2018.20.2/smarek|issn=1294-8322|pmc=6136121|pmid=30250390}}</ref><br />
<br />
*Versions of this network have also been called the central executive (or executive control) network and the cognitive control network.<ref name="Uddin2019" /><br />
<br />
<br />
===控制网络（侧额顶骨）===<br />
=<nowiki>= = 控制网络(侧额顶骨) = =</nowiki>= <br />
*这个网络参与了认知控制的启动与调节，它包括了大脑的18个亚区。<ref name=":14" />额顶网络与其他网络的参与程度和流体智力之间存在着有很强的相关性。<ref name=":15" /><br />
*在其它版本中，这种网络也被称为中央执行(或执行控制)网络和认知控制网络。<ref name="Uddin2019" /><br />
<br />
===Sensorimotor or Somatomotor (Pericentral)===<br />
{{Main|Sensorimotor network}}<br />
*This network processes somatosensory information and coordinates motion.<ref name="Heine" /><ref name="Yeo" /><ref name="Shafiei" /><ref name="Bassett" /><ref name="Yuan" /> The [[auditory cortex]] may be included.<ref name="Uddin2019" /><ref name="Yeo" /><br />
<br />
===感觉运动网络（中央区域）===<br />
=<nowiki>= = 感觉运动网络(中央区域) = =</nowiki>= <br />
*这个网络参与了躯体感觉信息的加工和运动的协调。<ref name="Heine" /><ref name="Yeo" /><ref name="Shafiei" /><ref name="Bassett" /><ref name="Yuan" />'''<font color="#ff8000">听觉皮层Auditory cortex</font>'''可能也包括在内。<ref name="Uddin2019" /><ref name="Yeo" /><br />
<br />
===Visual (Occipital)===<br />
{{See|Visual cortex}}<br />
*This network handles visual information processing.<ref name="Yang">{{cite journal|last1=Yang|first1=Yan-li|last2=Deng|first2=Hong-xia|last3=Xing|first3=Gui-yang|last4=Xia|first4=Xiao-luan|last5=Li|first5=Hai-fang|title=Brain functional network connectivity based on a visual task: visual information processing-related brain regions are significantly activated in the task state|journal=Neural Regeneration Research|date=2015|volume=10|issue=2|pages=298–307|doi=10.4103/1673-5374.152386|pmid=25883631|pmc=4392680 }}</ref><br />
<br />
===视觉网络（枕部）===<br />
=<nowiki>= = 视觉网络(枕部) = =</nowiki>= <br />
*这个网络处理视觉信息。<ref name="Yang" /><br />
<br />
==Other networks==<br />
Different methods and data have identified several other brain networks, many of which greatly overlap or are subsets of more well-characterized core networks.<ref name="Uddin2019" /><br />
*Limbic<ref name="Bassett" /><ref name="Yeo" /><ref name="Bailey" /><br />
*Auditory<ref name="Yuan" /><ref name="Heine" /><br />
*Right/left executive<ref name="Yuan" /><ref name="Heine" /><br />
*Cerebellar<ref name="Bell" /><ref name="Heine" /><br />
*Spatial attention<ref name="Riedl" /><ref name="Bressler" /><br />
*Language<ref name="Bressler" /><ref name="Hutton" /><br />
*Lateral visual<ref name="Yuan" /><ref name="Bell" /><ref name="Heine" /><br />
*Temporal<ref name="Yeo" /><ref name="Shafiei" /><br />
*Visual perception/imagery<ref name="Hutton" /><br />
<br />
==其他网络==<br />
=<nowiki>= = 其他网络 = =</nowiki>= <br />
不同的方法和数据已经能够确定上述核心网络以外的几个大脑网络，其中许多网络之间存在极大的重叠，部分网络实际上是核心网络更具特点的子集。<ref name="Uddin2019" /><br />
*边缘网络<ref name="Bassett" /><ref name="Yeo" /><ref name="Bailey" /><br />
*听觉网络<ref name="Yuan" /><ref name="Heine" /><br />
*右/左执行网络<ref name="Yuan" /><ref name="Heine" /><br />
*小脑网络<ref name="Bell" /><ref name="Heine" /><br />
*空间注意网络<ref name="Riedl" /><ref name="Bressler" /><br />
*语言网络<ref name="Bressler" /><ref name="Hutton" /><br />
*外侧视觉网络<ref name="Yuan" /><ref name="Bell" /><ref name="Heine" /><br />
*颞骨网络<ref name="Yeo" /><ref name="Shafiei" /><br />
*颞视知觉/图像网络<ref name="Hutton" /><br />
<br />
==See also==<br />
*[[Complex network]]<br />
*[[Neural network]]<br />
<br />
==相关词条==<br />
=<nowiki>= = 相关词条 = =</nowiki>= <br />
*[[复杂网络]]<br />
*[[神经网络]]<br />
<br />
==References==<br />
{{reflist|30em}}<br />
<br />
{{Human connectomics}}<br />
<br />
[[Category:Neuroscience]]<br />
[[Category:Neural coding]]<br />
[[Category:Neural circuits]]<br />
[[Category:Neurophysiology]]<br />
<br />
Category:Neuroscience<br />
Category:Neural coding<br />
Category:Neural circuits<br />
Category:Neurophysiology<br />
<br />
类别: 神经科学类别: 神经编码类别: 神经回路类别: 神经生理学<br />
<br />
<noinclude><br />
<br />
<small>This page was moved from [[wikipedia:en:Large-scale brain network]]. Its edit history can be viewed at [[大规模脑网络/edithistory]]</small></noinclude><br />
<br />
[[Category:待整理页面]]</div>Shenky20https://wiki.swarma.org/index.php?title=%E5%A4%A7%E8%A7%84%E6%A8%A1%E8%84%91%E7%BD%91%E7%BB%9C&diff=29992大规模脑网络2022-04-04T09:37:56Z<p>Shenky20：</p>
<hr />
<div>此词条由神经动力学模型读书会词条梳理志愿者Shenky20翻译审校，翻译字数共800，未经专家审核，带来阅读不便，请见谅。<br />
<br />
{{Short description|Collections of brain regions}}<br />
'''Large-scale brain networks''' (also known as '''intrinsic brain networks''') are collections of widespread [[brain regions]] showing [[Resting state fMRI#Functional|functional connectivity]] by statistical analysis of the [[Functional magnetic resonance imaging|fMRI]] [[BOLD signal]]<ref name=Riedl>{{cite journal|last1=Riedl|first1=Valentin|last2=Utz|first2=Lukas|last3=Castrillón|first3=Gabriel|last4=Grimmer|first4=Timo|last5=Rauschecker|first5=Josef P.|last6=Ploner|first6=Markus|last7=Friston|first7=Karl J.|last8=Drzezga|first8=Alexander|last9=Sorg|first9=Christian|title=Metabolic connectivity mapping reveals effective connectivity in the resting human brain|journal=PNAS|date=January 12, 2016|volume=113|issue=2|pages=428–433|doi=10.1073/pnas.1513752113|pmid=26712010|pmc=4720331|bibcode=2016PNAS..113..428R|doi-access=free}}</ref> or other recording methods such as [[Electroencephalography|EEG]],<ref name=":1">{{Cite journal|last1=Foster|first1=Brett L.|last2=Parvizi|first2=Josef|date=2012-03-01|title=Resting oscillations and cross-frequency coupling in the human posteromedial cortex|journal=NeuroImage|volume=60|issue=1|pages=384–391|doi=10.1016/j.neuroimage.2011.12.019|pmid=22227048|issn=1053-8119|pmc=3596417}}</ref> [[Positron emission tomography|PET]]<ref name=":2">{{Cite journal|last1=Buckner|first1=Randy L.|last2=Andrews‐Hanna|first2=Jessica R.|last3=Schacter|first3=Daniel L.|date=2008|title=The Brain's Default Network|journal=Annals of the New York Academy of Sciences|language=en|volume=1124|issue=1|pages=1–38|doi=10.1196/annals.1440.011|pmid=18400922|issn=1749-6632|bibcode=2008NYASA1124....1B|s2cid=3167595}}</ref> and [[Magnetoencephalography|MEG]].<ref name=":3">{{Cite journal|last1=Morris|first1=Peter G.|last2=Smith|first2=Stephen M.|last3=Barnes|first3=Gareth R.|last4=Stephenson|first4=Mary C.|last5=Hale|first5=Joanne R.|last6=Price|first6=Darren|last7=Luckhoo|first7=Henry|last8=Woolrich|first8=Mark|last9=Brookes|first9=Matthew J.|date=2011-10-04|title=Investigating the electrophysiological basis of resting state networks using magnetoencephalography|journal=Proceedings of the National Academy of Sciences|language=en|volume=108|issue=40|pages=16783–16788|doi=10.1073/pnas.1112685108|issn=0027-8424|pmid=21930901|pmc=3189080|bibcode=2011PNAS..10816783B|doi-access=free}}</ref> An emerging paradigm in neuroscience is that cognitive tasks are performed not by individual brain regions working in isolation but by networks consisting of several discrete brain regions that are said to be "functionally connected". Functional connectivity networks may be found using algorithms such as [[cluster analysis]], spatial [[independent component analysis]] (ICA), seed based, and others.<ref name="Petersen">{{cite journal|last1=Petersen|first1=Steven|last2=Sporns|first2=Olaf|title=Brain Networks and Cognitive Architectures|journal=Neuron|date=October 2015|volume=88|issue=1|pages=207–219|doi=10.1016/j.neuron.2015.09.027|pmid=26447582|pmc=4598639 }}</ref> Synchronized brain regions may also be identified using long-range synchronization of the EEG, MEG, or other dynamic brain signals.<ref name=Bressler>{{cite journal|last1=Bressler|first1=Steven L.|last2=Menon|first2=Vinod|s2cid=5967761|title=Large scale brain networks in cognition: emerging methods and principles|journal=Trends in Cognitive Sciences|date=June 2010|volume=14|issue=6|pages=233–290|doi=10.1016/j.tics.2010.04.004|url=http://www.cell.com/trends/cognitive-sciences/issue?pii=S1364-6613(10)X0005-5|accessdate=24 January 2016|pmid=20493761}}</ref><br />
<br />
<br />
<br />
'''<font color="#ff8000">大规模脑网络Large-scale brain networks</font>'''(也称为'''<font color="#ff8000">内在大脑网络Intrinsic brain networks</font>''')是在对基于'''<font color="#ff8000">血氧水平依赖效应BOLD</font>'''的'''<font color="#ff8000">功能性磁共振成像fMRI</font>'''信号<ref name="Riedl" />的统计分析或其他记录方法(如'''<font color="#ff8000">脑电图EEG<ref name=":1" /></font>'''、'''<font color="#ff8000">正电子发射断层扫描技术PET<ref name=":2" /></font>'''和'''<font color="#ff8000">脑磁图MEG<ref name=":3" /></font>''')中，表现出'''<font color="#ff8000">功能连接Functional connectivity</font>'''的'''<font color="#ff8000">脑区Brain regions</font>'''的集合。根据神经科学中一个新出现的范式，认知任务不是由单个脑区独立执行的，而是由几个互不相连的脑区“功能连接”组成的网络执行的。功能连接网络可以通过'''<font color="#ff8000">数据聚类Cluster analysis</font>'''、空间'''<font color="#ff8000">独立元素分析ICA</font>'''、种子点方法等算法来发现。<ref name="Petersen" />同步的脑区也可以用脑电图、脑磁图或其他动态脑信号的远程同步来识别。<ref name="Bressler" /><br />
<br />
The set of identified brain areas that are linked together in a large-scale network varies with cognitive function.<ref name="Bressler2">{{cite journal|last1=Bressler|first1=Steven L.|title=Neurocognitive networks|journal=Scholarpedia|volume=3|issue=2|pages=1567|doi=10.4249/scholarpedia.1567|year=2008|bibcode=2008SchpJ...3.1567B|doi-access=free}}</ref> When the cognitive state is not explicit (i.e., the subject is at "rest"), the large-scale brain network is a [[Resting state fMRI|resting state]] network (RSN). As a physical system with graph-like properties,<ref name="Bressler" /> a large-scale brain network has both nodes and edges and cannot be identified simply by the co-activation of brain areas. In recent decades, the analysis of brain networks was made feasible by advances in imaging techniques as well as new tools from [[graph theory]] and [[Dynamical systems theory|dynamical systems]].<br />
<br />
<br />
大规模脑网络中,连接在一起的脑区集合因认知功能的不同而不同。<ref name="Bressler2" />当认知状态不明确(即主体处于“静止”状态)时，大规模脑网络是一个'''<font color="#ff8000">静息状态Resting state</font>'''网络(RSN)。作为一个具有图形特征的物理系统，<ref name="Bressler" />大规模脑网络既有节点又有边，不能简单地通过脑区的共同激活来识别。近几十年来，成像技术不断进步，同时'''<font color="#ff8000">图论Graph theory</font>'''、'''<font color="#ff8000">动力学系统Dynamical systems</font>'''领域出现了新的技术手段，这使得脑网络分析变得可行。<br />
<br />
Large-scale brain networks are identified by their function and provide a coherent framework for understanding [[cognition]] by offering a neural model of how different cognitive functions emerge when different sets of brain regions join together as self-organized coalitions. The number and composition of the coalitions will vary with the algorithm and parameters used to identify them.<ref name="Yeo" /><ref name=":4">{{cite journal|last1=Abou Elseoud|first1=Ahmed|last2=Littow|first2=Harri|last3=Remes|first3=Jukka|last4=Starck|first4=Tuomo|last5=Nikkinen|first5=Juha|last6=Nissilä|first6=Juuso|last7=Timonen|first7=Markku|last8=Tervonen|first8=Osmo|last9=Kiviniemi1|first9=Vesa|title=Group-ICA Model Order Highlights Patterns of Functional Brain Connectivity|journal= Frontiers in Systems Neuroscience|date=2011-06-03|volume=5|pages=37|doi=10.3389/fnsys.2011.00037|pmid=21687724|pmc=3109774 |doi-access=free}}</ref> In one model, there is only the [[default mode network]] and the [[task-positive network]], but most current analyses show several networks, from a small handful to 17.<ref name="Yeo" /> The most common and stable networks are enumerated below. The regions participating in a functional network may be dynamically reconfigured.<ref name="Petersen" /><ref name="Bassett">{{cite journal|last1=Bassett|first1=Daniella|last2=Bertolero|first2=Max|title=How Matter Becomes Mind|journal=Scientific American|date=July 2019|volume=321|issue=1|page=32|url=https://www.scientificamerican.com/|accessdate=23 June 2019 }}</ref><br />
<br />
<br />
大规模脑网络是通过其功能来进行识别的。通过研究大规模脑网络建立神经模型，对不同脑区组合所形成的自组织联合体如何实现不同的'''<font color="#ff8000">认知Cognition</font>'''功能进行解释，就能够为认知理解提供一个连贯的框架。识别算法和参数的不同会导致所识别出的上述联合体的数量和组成有所不同。<ref name="Yeo" /><ref name=":4" />一个模型理论认为，符合上述条件的神经模型只包含'''<font color="#ff8000">默认模式网络Default mode network</font>'''和'''<font color="#ff8000">任务激活网络Task-positive network</font>'''，但目前'''<font color="#32CD32">大多数分析理论都包括从几个到17个不等的网络。<ref name="Yeo" /></font>'''下面列举了最常见且稳定的网络。'''<font color="#32CD32"> 人脑</font>'''可以动态地重新配置参与功能网络的脑区。<ref name="Petersen" /><ref name="Bassett" /> <br />
<br />
Disruptions in activity in various networks have been implicated in neuropsychiatric disorders such as [[Depression (mood)|depression]], [[Alzheimer's disease|Alzheimer's]], [[Autism-spectrum disorder|autism spectrum disorder]], [[schizophrenia]], [[ADHD]]<ref name=":5">{{cite journal |last1=Griffiths |first1=Kristi R. |last2=Braund |first2=Taylor A. |last3=Kohn |first3=Michael R. |last4=Clarke |first4=Simon |last5=Williams |first5=Leanne M. |last6=Korgaonkar |first6=Mayuresh S. |title=Structural brain network topology underpinning ADHD and response to methylphenidate treatment |journal=Translational Psychiatry |date=2 March 2021 |volume=11 |issue=1 |pages=1–9 |doi=10.1038/s41398-021-01278-x | pmc=7925571 |pmid=33654073 |url=https://www.nature.com/articles/s41398-021-01278-x#citeas |access-date=16 November 2021}}</ref> and [[bipolar disorder]].<ref name=":6">{{Cite journal|url=https://www.researchgate.net/publication/51639686|title=Large-scale brain networks and psychopathology: A unifying triple network model|last=Menon|first=Vinod|s2cid=26653572|journal=Trends in Cognitive Sciences|date=2011-09-09|volume=15|issue=10|pages=483–506|doi=10.1016/j.tics.2011.08.003|pmid=21908230}}</ref><br />
<br />
<br />
脑网络活动的中断与诸多神经精神疾病密切相关，如'''<font color="#ff8000">抑郁症Depression</font>'''、'''<font color="#ff8000">老年痴呆症Alzheimer's</font>'''、'''<font color="#ff8000">自闭症谱系障碍Autism spectrum disorder</font>'''、'''<font color="#ff8000">精神分裂症Schizophrenia</font>'''、'''<font color="#ff8000">多动症ADHD<ref name=":5" /></font>'''和'''<font color="#ff8000">躁郁症Bipolar disorder<ref name=":6" /></font>'''。<br />
<br />
==Core networks==<br />
[[File:Heine2012x3010.png|thumb|An example that identified 10 large-scale brain networks from [[resting state fMRI]] activity through [[independent component analysis]].<ref name="Heine" />|链接=Special:FilePath/Heine2012x3010.png]]<br />
Because brain networks can be identified at various different resolutions and with various different neurobiological properties, there is no such thing as a universal atlas of brain networks that fits all circumstances.<ref name=":7">{{cite journal|last1=Eickhoff|first1=SB|last2=Yeo|first2=BTT|last3=Genon|first3=S|title=Imaging-based parcellations of the human brain.|journal=Nature Reviews. Neuroscience|date=November 2018|volume=19|issue=11|pages=672–686|doi=10.1038/s41583-018-0071-7|pmid=30305712|s2cid=52954265|url=http://juser.fz-juelich.de/record/856633/files/Eickhoff_Yeo_Genon_NRN_MainManuscriptInclFigures.pdf}}</ref> While acknowledging this problem, Uddin, Yeo, and Spreng proposed in 2019<ref name="Uddin2019">{{cite journal|last1=Uddin|first1=LQ|last2=Yeo|first2=BTT|last3=Spreng|first3=RN|title=Towards a Universal Taxonomy of Macro-scale Functional Human Brain Networks.|journal=Brain Topography|date=November 2019|volume=32|issue=6|pages=926–942|doi=10.1007/s10548-019-00744-6|pmid=31707621|pmc=7325607}}</ref> that the following six networks should be defined as core networks based on converging evidences from multiple studies<ref name=":8">{{cite journal|last1=Doucet|first1=GE|last2=Lee|first2=WH|last3=Frangou|first3=S|title=Evaluation of the spatial variability in the major resting-state networks across human brain functional atlases.|journal=Human Brain Mapping|date=2019-10-15|volume=40|issue=15|pages=4577–4587|doi=10.1002/hbm.24722|pmid=31322303|pmc=6771873}}</ref><ref name="Yeo" /><ref name=":9">{{cite journal|last1=Smith|first1=SM|last2=Fox|first2=PT|last3=Miller|first3=KL|last4=Glahn|first4=DC|last5=Fox|first5=PM|last6=Mackay|first6=CE|last7=Filippini|first7=N|last8=Watkins|first8=KE|last9=Toro|first9=R|last10=Laird|first10=AR|last11=Beckmann|first11=CF|title=Correspondence of the brain's functional architecture during activation and rest.|journal=Proceedings of the National Academy of Sciences of the United States of America|date=2009-08-04|volume=106|issue=31|pages=13040–5|doi=10.1073/pnas.0905267106|pmid=19620724|pmc=2722273|bibcode=2009PNAS..10613040S|doi-access=free}}</ref> to facilitate communication between researchers.<br />
<br />
<br />
=<nowiki>= = 核心网络 = =</nowiki>= <br />
【图1：An example that identified 10 large-scale brain networks from resting state fMRI activity through independent component analysis.<ref name="Heine" />这是一个通过独立元素分析方法从静息状态脑功能性磁共振成像信号中分辨出10个大规模脑网络的例子。<ref name="Heine" />】<br />
<br />
<br />
因为脑网络可以在不同分辨率条件下、使用不同的神经生物学特性来识别，所以没有能够适用于所有情况的通用脑网络图谱。<ref name=":7" />在承认这个问题的同时，Uddin，Yeo，和Spreng在2019年提出<ref name="Uddin2019" />，综合考虑来自多个研究的证据<ref name=":8" /><ref name="Yeo" /><ref name=":9" />，以下六个网络应该被定义为核心网络，以促进研究人员之间的交流。<br />
<br />
===Default Mode (Medial frontoparietal)===<br />
{{Main|Default mode network}}<br />
*The default mode network is active when an individual is awake and at rest. It preferentially activates when individuals focus on internally-oriented tasks such as daydreaming, envisioning the future, retrieving memories, and [[theory of mind]]. It is negatively correlated with brain systems that focus on external visual signals. It is the most widely researched network.<ref name="Bressler" /><ref name="Bassett" /><ref name=":10">{{Cite journal|date=2012-08-15|title=The serendipitous discovery of the brain's default network|journal=NeuroImage|language=en|volume=62|issue=2|pages=1137–1145|doi=10.1016/j.neuroimage.2011.10.035|pmid=22037421|issn=1053-8119|last1=Buckner|first1=Randy L.|s2cid=9880586}}</ref><ref name="Riedl" /><ref name="Yuan"><br />
{{cite journal|last1=Yuan|first1=Rui|last2=Di|first2=Xin|last3=Taylor|first3=Paul A.|last4=Gohel|first4=Suril|last5=Tsai|first5=Yuan-Hsiung|last6=Biswal|first6=Bharat B.|title=Functional topography of the thalamocortical system in human|journal=Brain Structure and Function|date=30 April 2015|doi=10.1007/s00429-015-1018-7|pmid=25924563|pmc=6363530|volume=221|issue=4|pages=1971–1984}}</ref><ref name="Bell">{{cite journal|last1=Bell|first1=Peter T.|last2=Shine|first2=James M.|title=Estimating Large-Scale Network Convergence in the Human Functional Connectome|journal=Brain Connectivity|date=2015-11-09|volume=5|issue=9|doi=10.1089/brain.2015.0348|pmid=26005099|pages=565–74}}</ref><ref name="Heine">{{cite journal|last1=Heine|first1=Lizette|last2=Soddu|first2=Andrea|last3=Gomez|first3=Francisco|last4=Vanhaudenhuyse|first4=Audrey|last5=Tshibanda|first5=Luaba|last6=Thonnard|first6=Marie|last7=Charland-Verville|first7=Vanessa|last8=Kirsch|first8=Murielle|last9=Laureys|first9=Steven|last10=Demertzi|first10=Athena|title=Resting state networks and consciousness. Alterations of multiple resting state network connectivity in physiological, pharmacological and pathological consciousness states.|journal=Frontiers in Psychology|date=2012|volume=3|pages=295|doi=10.3389/fpsyg.2012.00295|pmid=22969735|pmc=3427917|doi-access=free}}</ref><ref name="Yeo">{{cite journal|last1=Yeo|first1=B. T. Thomas|last2=Krienen|first2=Fenna M.|last3=Sepulcre|first3=Jorge|last4=Sabuncu|first4=Mert R.|last5=Lashkari|first5=Danial|last6=Hollinshead|first6=Marisa|last7=Roffman|first7=Joshua L.|last8=Smoller|first8=Jordan W.|last9=Zöllei|first9=Lilla|last10=Polimeni|first10=Jonathan R.|last11=Fischl|first11=Bruce|last12=Liu|first12=Hesheng|last13=Buckner|first13=Randy L.|title=The organization of the human cerebral cortex estimated by intrinsic functional connectivity|journal=Journal of Neurophysiology|date=2011-09-01|volume=106|issue=3|pages=1125–1165|doi=10.1152/jn.00338.2011|pmid=21653723|pmc=3174820|bibcode=2011NatSD...2E0031H }}</ref><ref name="Shafiei">{{cite journal|last1=Shafiei|first1=Golia|last2=Zeighami|first2=Yashar|last3=Clark|first3=Crystal A.|last4=Coull|first4=Jennifer T.|last5=Nagano-Saito|first5=Atsuko|last6=Leyton|first6=Marco|last7=Dagher|first7=Alain|last8=Mišić|first8=Bratislav|title=Dopamine Signaling Modulates the Stability and Integration of Intrinsic Brain Networks|journal=Cerebral Cortex|date=2018-10-01|volume=29|issue=1|pages=397–409|doi=10.1093/cercor/bhy264|pmid=30357316|pmc=6294404 }}</ref><ref name="Bailey">{{cite journal|last1=Bailey|first1=Stephen K.|last2=Aboud|first2=Katherine S.|last3=Nguyen|first3=Tin Q.|last4=Cutting|first4=Laurie E.|title=Applying a network framework to the neurobiology of reading and dyslexia|journal=Journal of Neurodevelopmental Disorders|date=13 December 2018|volume=10|issue=1|page=37|doi=10.1186/s11689-018-9251-z|pmid=30541433|pmc=6291929 }}</ref><br />
<br />
<br />
=<nowiki>= = 默认模式网络（内侧额顶叶） = =</nowiki>= <br />
*默认模式网络在个体清醒和休息时都处于活跃状态。当个体专注于内部导向任务，比如做白日梦、展望未来、提取回忆和'''<font color="#ff8000">心智理论Theory of mind</font>'''任务时，默认模式网络会被优先激活。它与专注于外部视觉信号的大脑系统成负相关。对默认模式网络的研究目前是所谓网络之中最为广泛的。<ref name="Bressler" /><ref name="Bassett" /><ref name=":10" /><ref name="Riedl" /><ref name="Yuan" /><ref name="Bell" /><ref name="Heine" /><ref name="Yeo" /><ref name="Shafiei" /><ref name="Bailey" /><br />
<br />
===Salience (Midcingulo-Insular)===<br />
{{Main|Salience network}}<br />
*The salience network consists of several structures, including the anterior (bilateral) insula, dorsal anterior cingulate cortex, and three subcortical structures which are the ventral striatum, substantia nigra/ventral tegmental region.<ref name=":11">{{Cite journal|last1=Steimke|first1=Rosa|last2=Nomi|first2=Jason S.|last3=Calhoun|first3=Vince D.|last4=Stelzel|first4=Christine|last5=Paschke|first5=Lena M.|last6=Gaschler|first6=Robert|last7=Goschke|first7=Thomas|last8=Walter|first8=Henrik|last9=Uddin|first9=Lucina Q.|date=2017-12-01|title=Salience network dynamics underlying successful resistance of temptation|journal=Social Cognitive and Affective Neuroscience|language=en|volume=12|issue=12|pages=1928–1939|doi=10.1093/scan/nsx123|pmid=29048582|pmc=5716209|issn=1749-5016|doi-access=free}}</ref><ref name=":0">{{Citation|last=Menon|first=V.|title=Brain Mapping|chapter=Salience Network|date=2015-01-01|chapter-url=http://www.sciencedirect.com/science/article/pii/B978012397025100052X|pages=597–611|editor-last=Toga|editor-first=Arthur W.|publisher=Academic Press|isbn=978-0-12-397316-0|access-date=2019-12-08|doi=10.1016/B978-0-12-397025-1.00052-X}}</ref> It plays the key role of monitoring the [[Salience (neuroscience)|salience]] of external inputs and internal brain events.<ref name="Riedl" /><ref name="Bressler" /><ref name="Bassett" /><ref name="Yuan" /><ref name="Heine" /><ref name="Yeo" /><ref name="Shafiei" /> Specifically, it aids in directing attention by identifying important biological and cognitive events.<ref name=":0" /><ref name="Bailey" /><br />
*This network includes the ventral attention network, which primarily includes the [[temporoparietal junction]] and the ventral [[frontal cortex]] of the right hemisphere.<ref name="Uddin2019" /><ref name="Vossel" /> These areas respond when behaviorally relevant stimuli occur unexpectedly.<ref name="Vossel" /> The ventral attention network is inhibited during focused attention in which top-down processing is being used, such as when visually searching for something. This response may prevent goal-driven attention from being distracted by non-relevant stimuli. It becomes active again when the target or relevant information about the target is found.<ref name="Vossel" /><ref name=":12">{{Cite journal|last1=Shulman|first1=Gordon L.|last2=McAvoy|first2=Mark P.|last3=Cowan|first3=Melanie C.|last4=Astafiev|first4=Serguei V.|last5=Tansy|first5=Aaron P.|last6=d'Avossa|first6=Giovanni|last7=Corbetta|first7=Maurizio|date=2003-11-01|title=Quantitative Analysis of Attention and Detection Signals During Visual Search|journal=Journal of Neurophysiology|volume=90|issue=5|pages=3384–3397|doi=10.1152/jn.00343.2003|pmid=12917383|issn=0022-3077}}</ref><br />
<br />
<br />
=<nowiki>= = 突显网络（扣带回-岛叶）= =</nowiki>= <br />
*突显网络由'''<font color="#32CD32"> 前(双)岛、前扣带回皮层和三个皮层下结构（腹侧纹状体、黑质/腹侧被盖区）</font>'''组成，<ref name=":11" /><ref name=":0" />它在监测外部输入和内部脑活动的'''<font color="#ff8000">凸显程度Salience</font>'''中发挥着关键作用。<ref name="Riedl" /><ref name="Bressler" /><ref name="Bassett" /><ref name="Yuan" /><ref name="Heine" /><ref name="Yeo" /><ref name="Shafiei" />具体来说，突显网络通过识别重要的生理、认知活动来帮助引导注意力。<ref name=":0" /><ref name="Bailey" /><br />
*突显网络包括主要由右半球的'''<font color="#ff8000">颞顶联合区Temporoparietal junction</font>'''和腹侧'''<font color="#ff8000">额叶Frontal cortex</font>'''皮层组成的腹侧注意网络。<ref name="Uddin2019" /><ref name="Vossel" />当行为相关的刺激意外发生时，这些区域会对此做出反应。在自上而下加工注意焦点的过程中（例如使用视觉搜索某件物品时），腹侧注意网络会受到抑制。这种抑制反应可以防止目标驱动的注意力被不相关的刺激分散。当找到目标或相关信息时，突显网络会被再次激活。<ref name="Vossel" /><ref name=":12" /><br />
<br />
===Attention (Dorsal frontoparietal)===<br />
{{Main|Dorsal attention network}}<br />
*This network is involved in the voluntary, top-down deployment of attention.<ref name="Riedl" /><ref name="Yuan" /><ref name="Bell" /><ref name="Yeo" /><ref name="Shafiei" /><ref name="Vossel">{{cite journal|last1=Vossel|first1=Simone|last2=Geng|first2=Joy J.|last3=Fink|first3=Gereon R.|title=Dorsal and Ventral Attention Systems: Distinct Neural Circuits but Collaborative Roles|journal=The Neuroscientist|date=2014|volume=20|issue=2|pages=150–159|doi=10.1177/1073858413494269|pmid=23835449|pmc=4107817}}</ref><ref name="Hutton">{{cite journal|last1=Hutton|first1=John S.|last2=Dudley|first2=Jonathan|last3=Horowitz-Kraus|first3=Tzipi|last4=DeWitt|first4=Tom|last5=Holland|first5=Scott K.|title=Functional Connectivity of Attention, Visual, and Language Networks During Audio, Illustrated, and Animated Stories in Preschool-Age Children|journal=Brain Connectivity|date=1 September 2019|volume=9|issue=7|pages=580–592|doi=10.1089/brain.2019.0679|pmid=31144523|pmc=6775495|ref=Hutton}}</ref> Within the dorsal attention network, the intraparietal sulcus and frontal eye fields influence the visual areas of the brain. These influencing factors allow for the orientation of attention.<ref name=":13">{{Cite journal|last1=Fox|first1=Michael D.|last2=Corbetta|first2=Maurizio|last3=Snyder|first3=Abraham Z.|last4=Vincent|first4=Justin L.|last5=Raichle|first5=Marcus E.|date=2006-06-27|title=Spontaneous neuronal activity distinguishes human dorsal and ventral attention systems|journal=Proceedings of the National Academy of Sciences|language=en|volume=103|issue=26|pages=10046–10051|doi=10.1073/pnas.0604187103|issn=0027-8424|pmid=16788060|pmc=1480402|bibcode=2006PNAS..10310046F|doi-access=free}}</ref><ref name="Vossel" /><ref name="Bailey" /><br />
<br />
<br />
=<nowiki>= = 注意网络(背侧额顶骨) = =</nowiki>= <br />
*这个网络参与了自发的、自上而下的注意力分配。<ref name="Riedl" /><ref name="Yuan" /><ref name="Bell" /><ref name="Yeo" /><ref name="Shafiei" /><ref name="Vossel" /><ref name="Hutton" />在背侧注意网络中，顶内沟和额眼影响着大脑的视觉区域。这些影响因素决定了注意力的方向。<ref name=":13" /><ref name="Vossel" /><ref name="Bailey" /><br />
<br />
===Control (Lateral frontoparietal)===<br />
{{Main|Frontoparietal network}}<br />
*This network initiates and modulates cognitive control and comprises 18 sub-regions of the brain.<ref name=":14">{{Cite journal|last1=Scolari|first1=Miranda|last2=Seidl-Rathkopf|first2=Katharina N|last3=Kastner|first3=Sabine|date=2015-02-01|title=Functions of the human frontoparietal attention network: Evidence from neuroimaging|journal=Current Opinion in Behavioral Sciences|series=Cognitive control|volume=1|pages=32–39|doi=10.1016/j.cobeha.2014.08.003|issn=2352-1546|pmid=27398396|pmc=4936532}}</ref> There is a strong correlation between fluid intelligence and the involvement of the fronto-parietal network with other networks.<ref name=":15">{{Cite journal|last1=Marek|first1=Scott|last2=Dosenbach|first2=Nico U. F.|date=June 2018|title=The frontoparietal network: function, electrophysiology, and importance of individual precision mapping|journal=Dialogues in Clinical Neuroscience|volume=20|issue=2|pages=133–140|doi=10.31887/DCNS.2018.20.2/smarek|issn=1294-8322|pmc=6136121|pmid=30250390}}</ref><br />
<br />
*Versions of this network have also been called the central executive (or executive control) network and the cognitive control network.<ref name="Uddin2019" /><br />
<br />
<br />
=<nowiki>= = 控制网络(侧额顶骨) = =</nowiki>= <br />
*这个网络参与了认知控制的启动与调节，它包括了大脑的18个亚区。<ref name=":14" />额顶网络与其他网络的参与程度和流体智力之间存在着有很强的相关性。<ref name=":15" /><br />
*在其它版本中，这种网络也被称为中央执行(或执行控制)网络和认知控制网络。<ref name="Uddin2019" /><br />
<br />
===Sensorimotor or Somatomotor (Pericentral)===<br />
{{Main|Sensorimotor network}}<br />
*This network processes somatosensory information and coordinates motion.<ref name="Heine" /><ref name="Yeo" /><ref name="Shafiei" /><ref name="Bassett" /><ref name="Yuan" /> The [[auditory cortex]] may be included.<ref name="Uddin2019" /><ref name="Yeo" /><br />
<br />
=<nowiki>= = 感觉运动网络(中央区域) = =</nowiki>= <br />
*这个网络参与了躯体感觉信息的加工和运动的协调。<ref name="Heine" /><ref name="Yeo" /><ref name="Shafiei" /><ref name="Bassett" /><ref name="Yuan" />'''<font color="#ff8000">听觉皮层Auditory cortex</font>'''可能也包括在内。<ref name="Uddin2019" /><ref name="Yeo" /><br />
<br />
===Visual (Occipital)===<br />
{{See|Visual cortex}}<br />
*This network handles visual information processing.<ref name="Yang">{{cite journal|last1=Yang|first1=Yan-li|last2=Deng|first2=Hong-xia|last3=Xing|first3=Gui-yang|last4=Xia|first4=Xiao-luan|last5=Li|first5=Hai-fang|title=Brain functional network connectivity based on a visual task: visual information processing-related brain regions are significantly activated in the task state|journal=Neural Regeneration Research|date=2015|volume=10|issue=2|pages=298–307|doi=10.4103/1673-5374.152386|pmid=25883631|pmc=4392680 }}</ref><br />
<br />
=<nowiki>= = 视觉网络(枕部) = =</nowiki>= <br />
*这个网络处理视觉信息。<ref name="Yang" /><br />
<br />
==Other networks==<br />
Different methods and data have identified several other brain networks, many of which greatly overlap or are subsets of more well-characterized core networks.<ref name="Uddin2019" /><br />
* Limbic<ref name="Bassett" /><ref name="Yeo" /><ref name="Bailey" /><br />
*Auditory<ref name="Yuan" /><ref name="Heine" /><br />
*Right/left executive<ref name="Yuan" /><ref name="Heine" /><br />
*Cerebellar<ref name="Bell" /><ref name="Heine" /><br />
*Spatial attention<ref name="Riedl" /><ref name="Bressler" /><br />
*Language<ref name="Bressler" /><ref name="Hutton" /><br />
*Lateral visual<ref name="Yuan" /><ref name="Bell" /><ref name="Heine" /><br />
*Temporal<ref name="Yeo" /><ref name="Shafiei" /><br />
*Visual perception/imagery<ref name="Hutton" /><br />
<br />
=<nowiki>= = 其他网络 = =</nowiki>= <br />
不同的方法和数据已经能够确定上述核心网络以外的几个大脑网络，其中许多网络之间存在极大的重叠，部分网络实际上是核心网络更具特点的子集。<ref name="Uddin2019" /><br />
*边缘网络<ref name="Bassett" /><ref name="Yeo" /><ref name="Bailey" /><br />
*听觉网络<ref name="Yuan" /><ref name="Heine" /><br />
*右/左执行网络<ref name="Yuan" /><ref name="Heine" /><br />
* 小脑网络<ref name="Bell" /><ref name="Heine" /><br />
*空间注意网络<ref name="Riedl" /><ref name="Bressler" /><br />
*语言网络<ref name="Bressler" /><ref name="Hutton" /><br />
*外侧视觉网络<ref name="Yuan" /><ref name="Bell" /><ref name="Heine" /><br />
*颞骨网络<ref name="Yeo" /><ref name="Shafiei" /><br />
*颞视知觉/图像网络<ref name="Hutton" /><br />
<br />
==See also==<br />
*[[Complex network]]<br />
*[[Neural network]]<br />
<br />
=<nowiki>= = 相关词条 = =</nowiki>= <br />
*[[复杂网络]]<br />
*[[神经网络]]<br />
<br />
==References==<br />
{{reflist|30em}}<br />
<br />
{{Human connectomics}}<br />
<br />
[[Category:Neuroscience]]<br />
[[Category:Neural coding]]<br />
[[Category:Neural circuits]]<br />
[[Category:Neurophysiology]]<br />
<br />
Category:Neuroscience<br />
Category:Neural coding<br />
Category:Neural circuits<br />
Category:Neurophysiology<br />
<br />
类别: 神经科学类别: 神经编码类别: 神经回路类别: 神经生理学<br />
<br />
<noinclude><br />
<br />
<small>This page was moved from [[wikipedia:en:Large-scale brain network]]. Its edit history can be viewed at [[大规模脑网络/edithistory]]</small></noinclude><br />
<br />
[[Category:待整理页面]]</div>Shenky20https://wiki.swarma.org/index.php?title=%E5%A4%A7%E8%A7%84%E6%A8%A1%E8%84%91%E7%BD%91%E7%BB%9C&diff=29991大规模脑网络2022-04-04T09:34:49Z<p>Shenky20：</p>
<hr />
<div>此词条由神经动力学模型读书会词条梳理志愿者Shenky20翻译审校，翻译字数共800，未经专家审核，带来阅读不便，请见谅。<br />
<br />
{{Short description|Collections of brain regions}}<br />
'''Large-scale brain networks''' (also known as '''intrinsic brain networks''') are collections of widespread [[brain regions]] showing [[Resting state fMRI#Functional|functional connectivity]] by statistical analysis of the [[Functional magnetic resonance imaging|fMRI]] [[BOLD signal]]<ref name=Riedl>{{cite journal|last1=Riedl|first1=Valentin|last2=Utz|first2=Lukas|last3=Castrillón|first3=Gabriel|last4=Grimmer|first4=Timo|last5=Rauschecker|first5=Josef P.|last6=Ploner|first6=Markus|last7=Friston|first7=Karl J.|last8=Drzezga|first8=Alexander|last9=Sorg|first9=Christian|title=Metabolic connectivity mapping reveals effective connectivity in the resting human brain|journal=PNAS|date=January 12, 2016|volume=113|issue=2|pages=428–433|doi=10.1073/pnas.1513752113|pmid=26712010|pmc=4720331|bibcode=2016PNAS..113..428R|doi-access=free}}</ref> or other recording methods such as [[Electroencephalography|EEG]],<ref name=":1">{{Cite journal|last1=Foster|first1=Brett L.|last2=Parvizi|first2=Josef|date=2012-03-01|title=Resting oscillations and cross-frequency coupling in the human posteromedial cortex|journal=NeuroImage|volume=60|issue=1|pages=384–391|doi=10.1016/j.neuroimage.2011.12.019|pmid=22227048|issn=1053-8119|pmc=3596417}}</ref> [[Positron emission tomography|PET]]<ref name=":2">{{Cite journal|last1=Buckner|first1=Randy L.|last2=Andrews‐Hanna|first2=Jessica R.|last3=Schacter|first3=Daniel L.|date=2008|title=The Brain's Default Network|journal=Annals of the New York Academy of Sciences|language=en|volume=1124|issue=1|pages=1–38|doi=10.1196/annals.1440.011|pmid=18400922|issn=1749-6632|bibcode=2008NYASA1124....1B|s2cid=3167595}}</ref> and [[Magnetoencephalography|MEG]].<ref name=":3">{{Cite journal|last1=Morris|first1=Peter G.|last2=Smith|first2=Stephen M.|last3=Barnes|first3=Gareth R.|last4=Stephenson|first4=Mary C.|last5=Hale|first5=Joanne R.|last6=Price|first6=Darren|last7=Luckhoo|first7=Henry|last8=Woolrich|first8=Mark|last9=Brookes|first9=Matthew J.|date=2011-10-04|title=Investigating the electrophysiological basis of resting state networks using magnetoencephalography|journal=Proceedings of the National Academy of Sciences|language=en|volume=108|issue=40|pages=16783–16788|doi=10.1073/pnas.1112685108|issn=0027-8424|pmid=21930901|pmc=3189080|bibcode=2011PNAS..10816783B|doi-access=free}}</ref> An emerging paradigm in neuroscience is that cognitive tasks are performed not by individual brain regions working in isolation but by networks consisting of several discrete brain regions that are said to be "functionally connected". Functional connectivity networks may be found using algorithms such as [[cluster analysis]], spatial [[independent component analysis]] (ICA), seed based, and others.<ref name="Petersen">{{cite journal|last1=Petersen|first1=Steven|last2=Sporns|first2=Olaf|title=Brain Networks and Cognitive Architectures|journal=Neuron|date=October 2015|volume=88|issue=1|pages=207–219|doi=10.1016/j.neuron.2015.09.027|pmid=26447582|pmc=4598639 }}</ref> Synchronized brain regions may also be identified using long-range synchronization of the EEG, MEG, or other dynamic brain signals.<ref name=Bressler>{{cite journal|last1=Bressler|first1=Steven L.|last2=Menon|first2=Vinod|s2cid=5967761|title=Large scale brain networks in cognition: emerging methods and principles|journal=Trends in Cognitive Sciences|date=June 2010|volume=14|issue=6|pages=233–290|doi=10.1016/j.tics.2010.04.004|url=http://www.cell.com/trends/cognitive-sciences/issue?pii=S1364-6613(10)X0005-5|accessdate=24 January 2016|pmid=20493761}}</ref><br />
<br />
<br />
'''<font color="#ff8000">大规模脑网络Large-scale brain networks</font>'''(也称为'''<font color="#ff8000">内在大脑网络Intrinsic brain networks</font>''')是在对基于'''<font color="#ff8000">血氧水平依赖效应BOLD</font>'''的'''<font color="#ff8000">功能性磁共振成像fMRI</font>'''信号<ref name="Riedl" />的统计分析或其他记录方法(如'''<font color="#ff8000">脑电图EEG<ref name=":1" /></font>'''、'''<font color="#ff8000">正电子发射断层扫描技术PET<ref name=":2" /></font>'''和'''<font color="#ff8000">脑磁图MEG<ref name=":3" /></font>''')中，表现出'''<font color="#ff8000">功能连接Functional connectivity</font>'''的'''<font color="#ff8000">脑区Brain regions</font>'''的集合。根据神经科学中一个新出现的范式，认知任务不是由单个脑区独立执行的，而是由几个互不相连的脑区“功能连接”组成的网络执行的。功能连接网络可以通过'''<font color="#ff8000">数据聚类Cluster analysis</font>'''、空间'''<font color="#ff8000">独立元素分析ICA</font>'''、种子点方法等算法来发现。<ref name="Petersen" />同步的脑区也可以用脑电图、脑磁图或其他动态脑信号的远程同步来识别。<ref name="Bressler" /><br />
<br />
The set of identified brain areas that are linked together in a large-scale network varies with cognitive function.<ref name="Bressler2">{{cite journal|last1=Bressler|first1=Steven L.|title=Neurocognitive networks|journal=Scholarpedia|volume=3|issue=2|pages=1567|doi=10.4249/scholarpedia.1567|year=2008|bibcode=2008SchpJ...3.1567B|doi-access=free}}</ref> When the cognitive state is not explicit (i.e., the subject is at "rest"), the large-scale brain network is a [[Resting state fMRI|resting state]] network (RSN). As a physical system with graph-like properties,<ref name="Bressler" /> a large-scale brain network has both nodes and edges and cannot be identified simply by the co-activation of brain areas. In recent decades, the analysis of brain networks was made feasible by advances in imaging techniques as well as new tools from [[graph theory]] and [[Dynamical systems theory|dynamical systems]].<br />
<br />
<br />
大规模脑网络中,连接在一起的脑区集合因认知功能的不同而不同。<ref name="Bressler2" />当认知状态不明确(即主体处于“静止”状态)时，大规模脑网络是一个'''<font color="#ff8000">静息状态Resting state</font>'''网络(RSN)。作为一个具有图形特征的物理系统，<ref name="Bressler" />大规模脑网络既有节点又有边，不能简单地通过脑区的共同激活来识别。近几十年来，成像技术不断进步，同时'''<font color="#ff8000">图论Graph theory</font>'''、'''<font color="#ff8000">动力学系统Dynamical systems</font>'''领域出现了新的技术手段，这使得脑网络分析变得可行。<br />
<br />
Large-scale brain networks are identified by their function and provide a coherent framework for understanding [[cognition]] by offering a neural model of how different cognitive functions emerge when different sets of brain regions join together as self-organized coalitions. The number and composition of the coalitions will vary with the algorithm and parameters used to identify them.<ref name="Yeo" /><ref name=":4">{{cite journal|last1=Abou Elseoud|first1=Ahmed|last2=Littow|first2=Harri|last3=Remes|first3=Jukka|last4=Starck|first4=Tuomo|last5=Nikkinen|first5=Juha|last6=Nissilä|first6=Juuso|last7=Timonen|first7=Markku|last8=Tervonen|first8=Osmo|last9=Kiviniemi1|first9=Vesa|title=Group-ICA Model Order Highlights Patterns of Functional Brain Connectivity|journal= Frontiers in Systems Neuroscience|date=2011-06-03|volume=5|pages=37|doi=10.3389/fnsys.2011.00037|pmid=21687724|pmc=3109774 |doi-access=free}}</ref> In one model, there is only the [[default mode network]] and the [[task-positive network]], but most current analyses show several networks, from a small handful to 17.<ref name="Yeo" /> The most common and stable networks are enumerated below. The regions participating in a functional network may be dynamically reconfigured.<ref name="Petersen" /><ref name="Bassett">{{cite journal|last1=Bassett|first1=Daniella|last2=Bertolero|first2=Max|title=How Matter Becomes Mind|journal=Scientific American|date=July 2019|volume=321|issue=1|page=32|url=https://www.scientificamerican.com/|accessdate=23 June 2019 }}</ref><br />
<br />
<br />
大规模脑网络是通过其功能来进行识别的。通过研究大规模脑网络建立神经模型，对不同脑区组合所形成的自组织联合体如何实现不同的'''<font color="#ff8000">认知Cognition</font>'''功能进行解释，就能够为认知理解提供一个连贯的框架。识别算法和参数的不同会导致所识别出的上述联合体的数量和组成有所不同。<ref name="Yeo" /><ref name=":4" />一个模型理论认为，符合上述条件的神经模型只包含'''<font color="#ff8000">默认模式网络Default mode network</font>'''和'''<font color="#ff8000">任务激活网络Task-positive network</font>'''，但目前'''<font color="#32CD32">大多数分析理论都包括从几个到17个不等的网络。<ref name="Yeo" /></font>'''下面列举了最常见且稳定的网络。'''<font color="#32CD32"> 人脑</font>'''可以动态地重新配置参与功能网络的脑区。<ref name="Petersen" /><ref name="Bassett" /> <br />
<br />
Disruptions in activity in various networks have been implicated in neuropsychiatric disorders such as [[Depression (mood)|depression]], [[Alzheimer's disease|Alzheimer's]], [[Autism-spectrum disorder|autism spectrum disorder]], [[schizophrenia]], [[ADHD]]<ref name=":5">{{cite journal |last1=Griffiths |first1=Kristi R. |last2=Braund |first2=Taylor A. |last3=Kohn |first3=Michael R. |last4=Clarke |first4=Simon |last5=Williams |first5=Leanne M. |last6=Korgaonkar |first6=Mayuresh S. |title=Structural brain network topology underpinning ADHD and response to methylphenidate treatment |journal=Translational Psychiatry |date=2 March 2021 |volume=11 |issue=1 |pages=1–9 |doi=10.1038/s41398-021-01278-x | pmc=7925571 |pmid=33654073 |url=https://www.nature.com/articles/s41398-021-01278-x#citeas |access-date=16 November 2021}}</ref> and [[bipolar disorder]].<ref name=":6">{{Cite journal|url=https://www.researchgate.net/publication/51639686|title=Large-scale brain networks and psychopathology: A unifying triple network model|last=Menon|first=Vinod|s2cid=26653572|journal=Trends in Cognitive Sciences|date=2011-09-09|volume=15|issue=10|pages=483–506|doi=10.1016/j.tics.2011.08.003|pmid=21908230}}</ref><br />
<br />
<br />
脑网络活动的中断与诸多神经精神疾病密切相关，如'''<font color="#ff8000">抑郁症Depression</font>'''、'''<font color="#ff8000">老年痴呆症Alzheimer's</font>'''、'''<font color="#ff8000">自闭症谱系障碍Autism spectrum disorder</font>'''、'''<font color="#ff8000">精神分裂症Schizophrenia</font>'''、'''<font color="#ff8000">多动症ADHD<ref name=":5" /></font>'''和'''<font color="#ff8000">躁郁症Bipolar disorder<ref name=":6" /></font>'''。<br />
<br />
== Core networks==<br />
[[File:Heine2012x3010.png|thumb|An example that identified 10 large-scale brain networks from [[resting state fMRI]] activity through [[independent component analysis]].<ref name="Heine" />|链接=Special:FilePath/Heine2012x3010.png]]<br />
Because brain networks can be identified at various different resolutions and with various different neurobiological properties, there is no such thing as a universal atlas of brain networks that fits all circumstances.<ref name=":7">{{cite journal|last1=Eickhoff|first1=SB|last2=Yeo|first2=BTT|last3=Genon|first3=S|title=Imaging-based parcellations of the human brain.|journal=Nature Reviews. Neuroscience|date=November 2018|volume=19|issue=11|pages=672–686|doi=10.1038/s41583-018-0071-7|pmid=30305712|s2cid=52954265|url=http://juser.fz-juelich.de/record/856633/files/Eickhoff_Yeo_Genon_NRN_MainManuscriptInclFigures.pdf}}</ref> While acknowledging this problem, Uddin, Yeo, and Spreng proposed in 2019<ref name="Uddin2019">{{cite journal|last1=Uddin|first1=LQ|last2=Yeo|first2=BTT|last3=Spreng|first3=RN|title=Towards a Universal Taxonomy of Macro-scale Functional Human Brain Networks.|journal=Brain Topography|date=November 2019|volume=32|issue=6|pages=926–942|doi=10.1007/s10548-019-00744-6|pmid=31707621|pmc=7325607}}</ref> that the following six networks should be defined as core networks based on converging evidences from multiple studies<ref name=":8">{{cite journal|last1=Doucet|first1=GE|last2=Lee|first2=WH|last3=Frangou|first3=S|title=Evaluation of the spatial variability in the major resting-state networks across human brain functional atlases.|journal=Human Brain Mapping|date=2019-10-15|volume=40|issue=15|pages=4577–4587|doi=10.1002/hbm.24722|pmid=31322303|pmc=6771873}}</ref><ref name="Yeo" /><ref name=":9">{{cite journal|last1=Smith|first1=SM|last2=Fox|first2=PT|last3=Miller|first3=KL|last4=Glahn|first4=DC|last5=Fox|first5=PM|last6=Mackay|first6=CE|last7=Filippini|first7=N|last8=Watkins|first8=KE|last9=Toro|first9=R|last10=Laird|first10=AR|last11=Beckmann|first11=CF|title=Correspondence of the brain's functional architecture during activation and rest.|journal=Proceedings of the National Academy of Sciences of the United States of America|date=2009-08-04|volume=106|issue=31|pages=13040–5|doi=10.1073/pnas.0905267106|pmid=19620724|pmc=2722273|bibcode=2009PNAS..10613040S|doi-access=free}}</ref> to facilitate communication between researchers.<br />
<br />
【图1：An example that identified 10 large-scale brain networks from resting state fMRI activity through independent component analysis.<ref name="Heine" />这是一个通过独立元素分析方法从静息状态脑功能性磁共振成像信号中分辨出10个大规模脑网络的例子。<ref name="Heine" />】<br />
<br />
<br />
因为脑网络可以在不同分辨率条件下、使用不同的神经生物学特性来识别，所以没有能够适用于所有情况的通用脑网络图谱。<ref name=":7" />在承认这个问题的同时，Uddin，Yeo，和Spreng在2019年提出<ref name="Uddin2019" />，综合考虑来自多个研究的证据<ref name=":8" /><ref name="Yeo" /><ref name=":9" />，以下六个网络应该被定义为核心网络，以促进研究人员之间的交流。<br />
<br />
===Default Mode (Medial frontoparietal)===<br />
{{Main|Default mode network}}<br />
*The default mode network is active when an individual is awake and at rest. It preferentially activates when individuals focus on internally-oriented tasks such as daydreaming, envisioning the future, retrieving memories, and [[theory of mind]]. It is negatively correlated with brain systems that focus on external visual signals. It is the most widely researched network.<ref name="Bressler" /><ref name="Bassett" /><ref name=":10">{{Cite journal|date=2012-08-15|title=The serendipitous discovery of the brain's default network|journal=NeuroImage|language=en|volume=62|issue=2|pages=1137–1145|doi=10.1016/j.neuroimage.2011.10.035|pmid=22037421|issn=1053-8119|last1=Buckner|first1=Randy L.|s2cid=9880586}}</ref><ref name="Riedl" /><ref name="Yuan"><br />
{{cite journal|last1=Yuan|first1=Rui|last2=Di|first2=Xin|last3=Taylor|first3=Paul A.|last4=Gohel|first4=Suril|last5=Tsai|first5=Yuan-Hsiung|last6=Biswal|first6=Bharat B.|title=Functional topography of the thalamocortical system in human|journal=Brain Structure and Function|date=30 April 2015|doi=10.1007/s00429-015-1018-7|pmid=25924563|pmc=6363530|volume=221|issue=4|pages=1971–1984}}</ref><ref name="Bell">{{cite journal|last1=Bell|first1=Peter T.|last2=Shine|first2=James M.|title=Estimating Large-Scale Network Convergence in the Human Functional Connectome|journal=Brain Connectivity|date=2015-11-09|volume=5|issue=9|doi=10.1089/brain.2015.0348|pmid=26005099|pages=565–74}}</ref><ref name="Heine">{{cite journal|last1=Heine|first1=Lizette|last2=Soddu|first2=Andrea|last3=Gomez|first3=Francisco|last4=Vanhaudenhuyse|first4=Audrey|last5=Tshibanda|first5=Luaba|last6=Thonnard|first6=Marie|last7=Charland-Verville|first7=Vanessa|last8=Kirsch|first8=Murielle|last9=Laureys|first9=Steven|last10=Demertzi|first10=Athena|title=Resting state networks and consciousness. Alterations of multiple resting state network connectivity in physiological, pharmacological and pathological consciousness states.|journal=Frontiers in Psychology|date=2012|volume=3|pages=295|doi=10.3389/fpsyg.2012.00295|pmid=22969735|pmc=3427917|doi-access=free}}</ref><ref name="Yeo">{{cite journal|last1=Yeo|first1=B. T. Thomas|last2=Krienen|first2=Fenna M.|last3=Sepulcre|first3=Jorge|last4=Sabuncu|first4=Mert R.|last5=Lashkari|first5=Danial|last6=Hollinshead|first6=Marisa|last7=Roffman|first7=Joshua L.|last8=Smoller|first8=Jordan W.|last9=Zöllei|first9=Lilla|last10=Polimeni|first10=Jonathan R.|last11=Fischl|first11=Bruce|last12=Liu|first12=Hesheng|last13=Buckner|first13=Randy L.|title=The organization of the human cerebral cortex estimated by intrinsic functional connectivity|journal=Journal of Neurophysiology|date=2011-09-01|volume=106|issue=3|pages=1125–1165|doi=10.1152/jn.00338.2011|pmid=21653723|pmc=3174820|bibcode=2011NatSD...2E0031H }}</ref><ref name="Shafiei">{{cite journal|last1=Shafiei|first1=Golia|last2=Zeighami|first2=Yashar|last3=Clark|first3=Crystal A.|last4=Coull|first4=Jennifer T.|last5=Nagano-Saito|first5=Atsuko|last6=Leyton|first6=Marco|last7=Dagher|first7=Alain|last8=Mišić|first8=Bratislav|title=Dopamine Signaling Modulates the Stability and Integration of Intrinsic Brain Networks|journal=Cerebral Cortex|date=2018-10-01|volume=29|issue=1|pages=397–409|doi=10.1093/cercor/bhy264|pmid=30357316|pmc=6294404 }}</ref><ref name="Bailey">{{cite journal|last1=Bailey|first1=Stephen K.|last2=Aboud|first2=Katherine S.|last3=Nguyen|first3=Tin Q.|last4=Cutting|first4=Laurie E.|title=Applying a network framework to the neurobiology of reading and dyslexia|journal=Journal of Neurodevelopmental Disorders|date=13 December 2018|volume=10|issue=1|page=37|doi=10.1186/s11689-018-9251-z|pmid=30541433|pmc=6291929 }}</ref><br />
<br />
<br />
=<nowiki>= = 默认模式网络（内侧额顶叶） = =</nowiki>= <br />
*默认模式网络在个体清醒和休息时都处于活跃状态。当个体专注于内部导向任务，比如做白日梦、展望未来、提取回忆和'''<font color="#ff8000">心智理论Theory of mind</font>'''任务时，默认模式网络会被优先激活。它与专注于外部视觉信号的大脑系统成负相关。对默认模式网络的研究目前是所谓网络之中最为广泛的。<ref name="Bressler" /><ref name="Bassett" /><ref name=":10" /><ref name="Riedl" /><ref name="Yuan" /><ref name="Bell" /><ref name="Heine" /><ref name="Yeo" /><ref name="Shafiei" /><ref name="Bailey" /><br />
<br />
===Salience (Midcingulo-Insular)===<br />
{{Main|Salience network}}<br />
*The salience network consists of several structures, including the anterior (bilateral) insula, dorsal anterior cingulate cortex, and three subcortical structures which are the ventral striatum, substantia nigra/ventral tegmental region.<ref name=":11">{{Cite journal|last1=Steimke|first1=Rosa|last2=Nomi|first2=Jason S.|last3=Calhoun|first3=Vince D.|last4=Stelzel|first4=Christine|last5=Paschke|first5=Lena M.|last6=Gaschler|first6=Robert|last7=Goschke|first7=Thomas|last8=Walter|first8=Henrik|last9=Uddin|first9=Lucina Q.|date=2017-12-01|title=Salience network dynamics underlying successful resistance of temptation|journal=Social Cognitive and Affective Neuroscience|language=en|volume=12|issue=12|pages=1928–1939|doi=10.1093/scan/nsx123|pmid=29048582|pmc=5716209|issn=1749-5016|doi-access=free}}</ref><ref name=":0">{{Citation|last=Menon|first=V.|title=Brain Mapping|chapter=Salience Network|date=2015-01-01|chapter-url=http://www.sciencedirect.com/science/article/pii/B978012397025100052X|pages=597–611|editor-last=Toga|editor-first=Arthur W.|publisher=Academic Press|isbn=978-0-12-397316-0|access-date=2019-12-08|doi=10.1016/B978-0-12-397025-1.00052-X}}</ref> It plays the key role of monitoring the [[Salience (neuroscience)|salience]] of external inputs and internal brain events.<ref name="Riedl" /><ref name="Bressler" /><ref name="Bassett" /><ref name="Yuan" /><ref name="Heine" /><ref name="Yeo" /><ref name="Shafiei" /> Specifically, it aids in directing attention by identifying important biological and cognitive events.<ref name=":0" /><ref name="Bailey" /><br />
*This network includes the ventral attention network, which primarily includes the [[temporoparietal junction]] and the ventral [[frontal cortex]] of the right hemisphere.<ref name="Uddin2019" /><ref name="Vossel" /> These areas respond when behaviorally relevant stimuli occur unexpectedly.<ref name="Vossel" /> The ventral attention network is inhibited during focused attention in which top-down processing is being used, such as when visually searching for something. This response may prevent goal-driven attention from being distracted by non-relevant stimuli. It becomes active again when the target or relevant information about the target is found.<ref name="Vossel" /><ref name=":12">{{Cite journal|last1=Shulman|first1=Gordon L.|last2=McAvoy|first2=Mark P.|last3=Cowan|first3=Melanie C.|last4=Astafiev|first4=Serguei V.|last5=Tansy|first5=Aaron P.|last6=d'Avossa|first6=Giovanni|last7=Corbetta|first7=Maurizio|date=2003-11-01|title=Quantitative Analysis of Attention and Detection Signals During Visual Search|journal=Journal of Neurophysiology|volume=90|issue=5|pages=3384–3397|doi=10.1152/jn.00343.2003|pmid=12917383|issn=0022-3077}}</ref><br />
<br />
<br />
=<nowiki>= = 突显网络（扣带回-岛叶）= =</nowiki>= <br />
*突显网络由'''<font color="#32CD32"> 前(双)岛、前扣带回皮层和三个皮层下结构（腹侧纹状体、黑质/腹侧被盖区）</font>'''组成，<ref name=":11" /><ref name=":0" />它在监测外部输入和内部脑活动的'''<font color="#ff8000">凸显程度Salience</font>'''中发挥着关键作用。<ref name="Riedl" /><ref name="Bressler" /><ref name="Bassett" /><ref name="Yuan" /><ref name="Heine" /><ref name="Yeo" /><ref name="Shafiei" />具体来说，突显网络通过识别重要的生理、认知活动来帮助引导注意力。<ref name=":0" /><ref name="Bailey" /><br />
*突显网络包括主要由右半球的'''<font color="#ff8000">颞顶联合区Temporoparietal junction</font>'''和腹侧'''<font color="#ff8000">额叶Frontal cortex</font>'''皮层组成的腹侧注意网络。<ref name="Uddin2019" /><ref name="Vossel" />当行为相关的刺激意外发生时，这些区域会对此做出反应。在自上而下加工注意焦点的过程中（例如使用视觉搜索某件物品时），腹侧注意网络会受到抑制。这种抑制反应可以防止目标驱动的注意力被不相关的刺激分散。当找到目标或相关信息时，突显网络会被再次激活。<ref name="Vossel" /><ref name=":12" /><br />
<br />
===Attention (Dorsal frontoparietal)===<br />
{{Main|Dorsal attention network}}<br />
*This network is involved in the voluntary, top-down deployment of attention.<ref name="Riedl" /><ref name="Yuan" /><ref name="Bell" /><ref name="Yeo" /><ref name="Shafiei" /><ref name="Vossel">{{cite journal|last1=Vossel|first1=Simone|last2=Geng|first2=Joy J.|last3=Fink|first3=Gereon R.|title=Dorsal and Ventral Attention Systems: Distinct Neural Circuits but Collaborative Roles|journal=The Neuroscientist|date=2014|volume=20|issue=2|pages=150–159|doi=10.1177/1073858413494269|pmid=23835449|pmc=4107817}}</ref><ref name="Hutton">{{cite journal|last1=Hutton|first1=John S.|last2=Dudley|first2=Jonathan|last3=Horowitz-Kraus|first3=Tzipi|last4=DeWitt|first4=Tom|last5=Holland|first5=Scott K.|title=Functional Connectivity of Attention, Visual, and Language Networks During Audio, Illustrated, and Animated Stories in Preschool-Age Children|journal=Brain Connectivity|date=1 September 2019|volume=9|issue=7|pages=580–592|doi=10.1089/brain.2019.0679|pmid=31144523|pmc=6775495|ref=Hutton}}</ref> Within the dorsal attention network, the intraparietal sulcus and frontal eye fields influence the visual areas of the brain. These influencing factors allow for the orientation of attention.<ref name=":13">{{Cite journal|last1=Fox|first1=Michael D.|last2=Corbetta|first2=Maurizio|last3=Snyder|first3=Abraham Z.|last4=Vincent|first4=Justin L.|last5=Raichle|first5=Marcus E.|date=2006-06-27|title=Spontaneous neuronal activity distinguishes human dorsal and ventral attention systems|journal=Proceedings of the National Academy of Sciences|language=en|volume=103|issue=26|pages=10046–10051|doi=10.1073/pnas.0604187103|issn=0027-8424|pmid=16788060|pmc=1480402|bibcode=2006PNAS..10310046F|doi-access=free}}</ref><ref name="Vossel" /><ref name="Bailey" /><br />
<br />
<br />
=<nowiki>= = 注意网络(背侧额顶骨) = =</nowiki>= <br />
*这个网络参与了自发的、自上而下的注意力分配。<ref name="Riedl" /><ref name="Yuan" /><ref name="Bell" /><ref name="Yeo" /><ref name="Shafiei" /><ref name="Vossel" /><ref name="Hutton" />在背侧注意网络中，顶内沟和额眼影响着大脑的视觉区域。这些影响因素决定了注意力的方向。<ref name=":13" /><ref name="Vossel" /><ref name="Bailey" /><br />
<br />
===Control (Lateral frontoparietal)===<br />
{{Main|Frontoparietal network}}<br />
*This network initiates and modulates cognitive control and comprises 18 sub-regions of the brain.<ref name=":14">{{Cite journal|last1=Scolari|first1=Miranda|last2=Seidl-Rathkopf|first2=Katharina N|last3=Kastner|first3=Sabine|date=2015-02-01|title=Functions of the human frontoparietal attention network: Evidence from neuroimaging|journal=Current Opinion in Behavioral Sciences|series=Cognitive control|volume=1|pages=32–39|doi=10.1016/j.cobeha.2014.08.003|issn=2352-1546|pmid=27398396|pmc=4936532}}</ref> There is a strong correlation between fluid intelligence and the involvement of the fronto-parietal network with other networks.<ref name=":15">{{Cite journal|last1=Marek|first1=Scott|last2=Dosenbach|first2=Nico U. F.|date=June 2018|title=The frontoparietal network: function, electrophysiology, and importance of individual precision mapping|journal=Dialogues in Clinical Neuroscience|volume=20|issue=2|pages=133–140|doi=10.31887/DCNS.2018.20.2/smarek|issn=1294-8322|pmc=6136121|pmid=30250390}}</ref><br />
<br />
*Versions of this network have also been called the central executive (or executive control) network and the cognitive control network.<ref name="Uddin2019" /><br />
<br />
<br />
=<nowiki>= = 控制网络(侧额顶骨) = =</nowiki>= <br />
*这个网络参与了认知控制的启动与调节，它包括了大脑的18个亚区。<ref name=":14" />额顶网络与其他网络的参与程度和流体智力之间存在着有很强的相关性。<ref name=":15" /><br />
*在其它版本中，这种网络也被称为中央执行(或执行控制)网络和认知控制网络。<ref name="Uddin2019" /><br />
<br />
===Sensorimotor or Somatomotor (Pericentral)===<br />
{{Main|Sensorimotor network}}<br />
*This network processes somatosensory information and coordinates motion.<ref name="Heine" /><ref name="Yeo" /><ref name="Shafiei" /><ref name="Bassett" /><ref name="Yuan" /> The [[auditory cortex]] may be included.<ref name="Uddin2019" /><ref name="Yeo" /><br />
<br />
=<nowiki>= = 感觉运动网络(中央区域) = =</nowiki>= <br />
* 这个网络参与了躯体感觉信息的加工和运动的协调。<ref name="Heine" /><ref name="Yeo" /><ref name="Shafiei" /><ref name="Bassett" /><ref name="Yuan" />'''<font color="#ff8000">听觉皮层Auditory cortex</font>'''可能也包括在内。<ref name="Uddin2019" /><ref name="Yeo" /><br />
<br />
===Visual (Occipital)===<br />
{{See|Visual cortex}}<br />
*This network handles visual information processing.<ref name="Yang">{{cite journal|last1=Yang|first1=Yan-li|last2=Deng|first2=Hong-xia|last3=Xing|first3=Gui-yang|last4=Xia|first4=Xiao-luan|last5=Li|first5=Hai-fang|title=Brain functional network connectivity based on a visual task: visual information processing-related brain regions are significantly activated in the task state|journal=Neural Regeneration Research|date=2015|volume=10|issue=2|pages=298–307|doi=10.4103/1673-5374.152386|pmid=25883631|pmc=4392680 }}</ref><br />
<br />
=<nowiki>= = 视觉网络(枕部) = =</nowiki>= <br />
*这个网络处理视觉信息。<ref name="Yang" /><br />
<br />
==Other networks==<br />
Different methods and data have identified several other brain networks, many of which greatly overlap or are subsets of more well-characterized core networks.<ref name="Uddin2019" /><br />
*Limbic<ref name="Bassett" /><ref name="Yeo" /><ref name="Bailey" /><br />
* Auditory<ref name="Yuan" /><ref name="Heine" /><br />
*Right/left executive<ref name="Yuan" /><ref name="Heine" /><br />
*Cerebellar<ref name="Bell" /><ref name="Heine" /><br />
*Spatial attention<ref name="Riedl" /><ref name="Bressler" /><br />
*Language<ref name="Bressler" /><ref name="Hutton" /><br />
*Lateral visual<ref name="Yuan" /><ref name="Bell" /><ref name="Heine" /><br />
*Temporal<ref name="Yeo" /><ref name="Shafiei" /><br />
* Visual perception/imagery<ref name="Hutton" /><br />
<br />
=<nowiki>= = 其他网络 = =</nowiki>= <br />
不同的方法和数据已经能够确定上述核心网络以外的几个大脑网络，其中许多网络之间存在极大的重叠，部分网络实际上是核心网络更具特点的子集。<ref name="Uddin2019" /><br />
*边缘网络<ref name="Bassett" /><ref name="Yeo" /><ref name="Bailey" /><br />
*听觉网络<ref name="Yuan" /><ref name="Heine" /><br />
*右/左执行网络<ref name="Yuan" /><ref name="Heine" /><br />
*小脑网络<ref name="Bell" /><ref name="Heine" /><br />
*空间注意网络<ref name="Riedl" /><ref name="Bressler" /><br />
*语言网络<ref name="Bressler" /><ref name="Hutton" /><br />
*外侧视觉网络<ref name="Yuan" /><ref name="Bell" /><ref name="Heine" /><br />
*颞骨网络<ref name="Yeo" /><ref name="Shafiei" /><br />
*颞视知觉/图像网络<ref name="Hutton" /><br />
<br />
==See also==<br />
*[[Complex network]]<br />
*[[Neural network]]<br />
<br />
=<nowiki>= = 相关词条 = =</nowiki>= <br />
*复杂网络<br />
*神经网络<br />
<br />
==References==<br />
{{reflist|30em}}<br />
<br />
{{Human connectomics}}<br />
<br />
[[Category:Neuroscience]]<br />
[[Category:Neural coding]]<br />
[[Category:Neural circuits]]<br />
[[Category:Neurophysiology]]<br />
<br />
Category:Neuroscience<br />
Category:Neural coding<br />
Category:Neural circuits<br />
Category:Neurophysiology<br />
<br />
类别: 神经科学类别: 神经编码类别: 神经回路类别: 神经生理学<br />
<br />
<noinclude><br />
<br />
<small>This page was moved from [[wikipedia:en:Large-scale brain network]]. Its edit history can be viewed at [[大规模脑网络/edithistory]]</small></noinclude><br />
<br />
[[Category:待整理页面]]</div>Shenky20https://wiki.swarma.org/index.php?title=%E5%A4%A7%E8%A7%84%E6%A8%A1%E8%84%91%E7%BD%91%E7%BB%9C&diff=29990大规模脑网络2022-04-04T09:14:25Z<p>Shenky20：</p>
<hr />
<div>此词条由神经动力学模型读书会词条梳理志愿者Shenky20翻译审校，翻译字数共800，未经专家审核，带来阅读不便，请见谅。<br />
<br />
{{Short description|Collections of brain regions}}<br />
'''Large-scale brain networks''' (also known as '''intrinsic brain networks''') are collections of widespread [[brain regions]] showing [[Resting state fMRI#Functional|functional connectivity]] by statistical analysis of the [[Functional magnetic resonance imaging|fMRI]] [[BOLD signal]]<ref name=Riedl>{{cite journal|last1=Riedl|first1=Valentin|last2=Utz|first2=Lukas|last3=Castrillón|first3=Gabriel|last4=Grimmer|first4=Timo|last5=Rauschecker|first5=Josef P.|last6=Ploner|first6=Markus|last7=Friston|first7=Karl J.|last8=Drzezga|first8=Alexander|last9=Sorg|first9=Christian|title=Metabolic connectivity mapping reveals effective connectivity in the resting human brain|journal=PNAS|date=January 12, 2016|volume=113|issue=2|pages=428–433|doi=10.1073/pnas.1513752113|pmid=26712010|pmc=4720331|bibcode=2016PNAS..113..428R|doi-access=free}}</ref> or other recording methods such as [[Electroencephalography|EEG]],<ref name=":1">{{Cite journal|last1=Foster|first1=Brett L.|last2=Parvizi|first2=Josef|date=2012-03-01|title=Resting oscillations and cross-frequency coupling in the human posteromedial cortex|journal=NeuroImage|volume=60|issue=1|pages=384–391|doi=10.1016/j.neuroimage.2011.12.019|pmid=22227048|issn=1053-8119|pmc=3596417}}</ref> [[Positron emission tomography|PET]]<ref name=":2">{{Cite journal|last1=Buckner|first1=Randy L.|last2=Andrews‐Hanna|first2=Jessica R.|last3=Schacter|first3=Daniel L.|date=2008|title=The Brain's Default Network|journal=Annals of the New York Academy of Sciences|language=en|volume=1124|issue=1|pages=1–38|doi=10.1196/annals.1440.011|pmid=18400922|issn=1749-6632|bibcode=2008NYASA1124....1B|s2cid=3167595}}</ref> and [[Magnetoencephalography|MEG]].<ref name=":3">{{Cite journal|last1=Morris|first1=Peter G.|last2=Smith|first2=Stephen M.|last3=Barnes|first3=Gareth R.|last4=Stephenson|first4=Mary C.|last5=Hale|first5=Joanne R.|last6=Price|first6=Darren|last7=Luckhoo|first7=Henry|last8=Woolrich|first8=Mark|last9=Brookes|first9=Matthew J.|date=2011-10-04|title=Investigating the electrophysiological basis of resting state networks using magnetoencephalography|journal=Proceedings of the National Academy of Sciences|language=en|volume=108|issue=40|pages=16783–16788|doi=10.1073/pnas.1112685108|issn=0027-8424|pmid=21930901|pmc=3189080|bibcode=2011PNAS..10816783B|doi-access=free}}</ref> An emerging paradigm in neuroscience is that cognitive tasks are performed not by individual brain regions working in isolation but by networks consisting of several discrete brain regions that are said to be "functionally connected". Functional connectivity networks may be found using algorithms such as [[cluster analysis]], spatial [[independent component analysis]] (ICA), seed based, and others.<ref name="Petersen">{{cite journal|last1=Petersen|first1=Steven|last2=Sporns|first2=Olaf|title=Brain Networks and Cognitive Architectures|journal=Neuron|date=October 2015|volume=88|issue=1|pages=207–219|doi=10.1016/j.neuron.2015.09.027|pmid=26447582|pmc=4598639 }}</ref> Synchronized brain regions may also be identified using long-range synchronization of the EEG, MEG, or other dynamic brain signals.<ref name=Bressler>{{cite journal|last1=Bressler|first1=Steven L.|last2=Menon|first2=Vinod|s2cid=5967761|title=Large scale brain networks in cognition: emerging methods and principles|journal=Trends in Cognitive Sciences|date=June 2010|volume=14|issue=6|pages=233–290|doi=10.1016/j.tics.2010.04.004|url=http://www.cell.com/trends/cognitive-sciences/issue?pii=S1364-6613(10)X0005-5|accessdate=24 January 2016|pmid=20493761}}</ref><br />
<br />
<br />
Large-scale brain networks (also known as intrinsic brain networks) are collections of widespread brain regions showing functional connectivity by statistical analysis of the fMRI BOLD signal or other recording methods such as EEG, PET and MEG. An emerging paradigm in neuroscience is that cognitive tasks are performed not by individual brain regions working in isolation but by networks consisting of several discrete brain regions that are said to be "functionally connected". Functional connectivity networks may be found using algorithms such as cluster analysis, spatial independent component analysis (ICA), seed based, and others. Synchronized brain regions may also be identified using long-range synchronization of the EEG, MEG, or other dynamic brain signals.<br />
<br />
'''<font color="#ff8000">大规模脑网络Large-scale brain networks</font>'''(也称为'''<font color="#ff8000">内在大脑网络Intrinsic brain networks</font>''')是在对基于'''<font color="#ff8000">血氧水平依赖效应BOLD</font>'''的'''<font color="#ff8000">功能性磁共振成像fMRI</font>'''信号<ref name="Riedl" />的统计分析或其他记录方法(如'''<font color="#ff8000">脑电图EEG<ref name=":1" /></font>'''、'''<font color="#ff8000">正电子发射断层扫描技术PET<ref name=":2" /></font>'''和'''<font color="#ff8000">脑磁图MEG<ref name=":3" /></font>''')中，表现出'''<font color="#ff8000">功能连接Functional connectivity</font>'''的'''<font color="#ff8000">脑区Brain regions</font>'''的集合。根据神经科学中一个新出现的范式，认知任务不是由单个脑区独立执行的，而是由几个互不相连的脑区“功能连接”组成的网络执行的。功能连接网络可以通过'''<font color="#ff8000">数据聚类Cluster analysis</font>'''、空间'''<font color="#ff8000">独立元素分析ICA</font>'''、种子点方法等算法来发现。<ref name="Petersen" />同步的脑区也可以用脑电图、脑磁图或其他动态脑信号的远程同步来识别。<ref name="Bressler" /><br />
<br />
The set of identified brain areas that are linked together in a large-scale network varies with cognitive function.<ref name="Bressler2">{{cite journal|last1=Bressler|first1=Steven L.|title=Neurocognitive networks|journal=Scholarpedia|volume=3|issue=2|pages=1567|doi=10.4249/scholarpedia.1567|year=2008|bibcode=2008SchpJ...3.1567B|doi-access=free}}</ref> When the cognitive state is not explicit (i.e., the subject is at "rest"), the large-scale brain network is a [[Resting state fMRI|resting state]] network (RSN). As a physical system with graph-like properties,<ref name="Bressler" /> a large-scale brain network has both nodes and edges and cannot be identified simply by the co-activation of brain areas. In recent decades, the analysis of brain networks was made feasible by advances in imaging techniques as well as new tools from [[graph theory]] and [[Dynamical systems theory|dynamical systems]].<br />
<br />
The set of identified brain areas that are linked together in a large-scale network varies with cognitive function. When the cognitive state is not explicit (i.e., the subject is at "rest"), the large-scale brain network is a resting state network (RSN). As a physical system with graph-like properties, a large-scale brain network has both nodes and edges and cannot be identified simply by the co-activation of brain areas. In recent decades, the analysis of brain networks was made feasible by advances in imaging techniques as well as new tools from graph theory and dynamical systems.<br />
<br />
大规模脑网络中,连接在一起的脑区集合因认知功能的不同而不同。<ref name="Bressler2" />当认知状态不明确(即主体处于“静止”状态)时，大规模脑网络是一个'''<font color="#ff8000">静息状态Resting state</font>'''网络(RSN)。作为一个具有图形特征的物理系统，<ref name="Bressler" />大规模脑网络既有节点又有边，不能简单地通过脑区的共同激活来识别。近几十年来，成像技术不断进步，同时'''<font color="#ff8000">图论Graph theory</font>'''、'''<font color="#ff8000">动力学系统Dynamical systems</font>'''领域出现了新的技术手段，这使得脑网络分析变得可行。<br />
<br />
Large-scale brain networks are identified by their function and provide a coherent framework for understanding [[cognition]] by offering a neural model of how different cognitive functions emerge when different sets of brain regions join together as self-organized coalitions. The number and composition of the coalitions will vary with the algorithm and parameters used to identify them.<ref name="Yeo" /><ref name=":4">{{cite journal|last1=Abou Elseoud|first1=Ahmed|last2=Littow|first2=Harri|last3=Remes|first3=Jukka|last4=Starck|first4=Tuomo|last5=Nikkinen|first5=Juha|last6=Nissilä|first6=Juuso|last7=Timonen|first7=Markku|last8=Tervonen|first8=Osmo|last9=Kiviniemi1|first9=Vesa|title=Group-ICA Model Order Highlights Patterns of Functional Brain Connectivity|journal= Frontiers in Systems Neuroscience|date=2011-06-03|volume=5|pages=37|doi=10.3389/fnsys.2011.00037|pmid=21687724|pmc=3109774 |doi-access=free}}</ref> In one model, there is only the [[default mode network]] and the [[task-positive network]], but most current analyses show several networks, from a small handful to 17.<ref name="Yeo" /> The most common and stable networks are enumerated below. The regions participating in a functional network may be dynamically reconfigured.<ref name="Petersen" /><ref name="Bassett">{{cite journal|last1=Bassett|first1=Daniella|last2=Bertolero|first2=Max|title=How Matter Becomes Mind|journal=Scientific American|date=July 2019|volume=321|issue=1|page=32|url=https://www.scientificamerican.com/|accessdate=23 June 2019 }}</ref><br />
<br />
Large-scale brain networks are identified by their function and provide a coherent framework for understanding cognition by offering a neural model of how different cognitive functions emerge when different sets of brain regions join together as self-organized coalitions. The number and composition of the coalitions will vary with the algorithm and parameters used to identify them. In one model, there is only the default mode network and the task-positive network, but most current analyses show several networks, from a small handful to 17. The most common and stable networks are enumerated below. The regions participating in a functional network may be dynamically reconfigured.<br />
<br />
大规模脑网络是通过其功能来进行识别的。通过研究大规模脑网络建立神经模型，对不同脑区组合所形成的自组织联合体如何实现不同的'''<font color="#ff8000">认知Cognition</font>'''功能进行解释，就能够为认知理解提供一个连贯的框架。识别算法和参数的不同会导致所识别出的上述联合体的数量和组成有所不同。<ref name="Yeo" /><ref name=":4" />一个模型理论认为，符合上述条件的神经模型只包含'''<font color="#ff8000">默认模式网络Default mode network</font>'''和'''<font color="#ff8000">任务激活网络Task-positive network</font>'''，但目前'''<font color="#32CD32">大多数分析理论都包括从几个到17个不等的网络。<ref name="Yeo" /></font>'''下面列举了最常见且稳定的网络。'''<font color="#32CD32"> 人脑</font>'''可以动态地重新配置参与功能网络的脑区。<ref name="Petersen" /><ref name="Bassett" /> <br />
<br />
Disruptions in activity in various networks have been implicated in neuropsychiatric disorders such as [[Depression (mood)|depression]], [[Alzheimer's disease|Alzheimer's]], [[Autism-spectrum disorder|autism spectrum disorder]], [[schizophrenia]], [[ADHD]]<ref name=":5">{{cite journal |last1=Griffiths |first1=Kristi R. |last2=Braund |first2=Taylor A. |last3=Kohn |first3=Michael R. |last4=Clarke |first4=Simon |last5=Williams |first5=Leanne M. |last6=Korgaonkar |first6=Mayuresh S. |title=Structural brain network topology underpinning ADHD and response to methylphenidate treatment |journal=Translational Psychiatry |date=2 March 2021 |volume=11 |issue=1 |pages=1–9 |doi=10.1038/s41398-021-01278-x | pmc=7925571 |pmid=33654073 |url=https://www.nature.com/articles/s41398-021-01278-x#citeas |access-date=16 November 2021}}</ref> and [[bipolar disorder]].<ref name=":6">{{Cite journal|url=https://www.researchgate.net/publication/51639686|title=Large-scale brain networks and psychopathology: A unifying triple network model|last=Menon|first=Vinod|s2cid=26653572|journal=Trends in Cognitive Sciences|date=2011-09-09|volume=15|issue=10|pages=483–506|doi=10.1016/j.tics.2011.08.003|pmid=21908230}}</ref><br />
<br />
Disruptions in activity in various networks have been implicated in neuropsychiatric disorders such as depression, Alzheimer's, autism spectrum disorder, schizophrenia, ADHD and bipolar disorder.<br />
<br />
脑网络活动的中断与诸多神经精神疾病密切相关，如'''<font color="#ff8000">抑郁症Depression</font>'''、'''<font color="#ff8000">老年痴呆症Alzheimer's</font>'''、'''<font color="#ff8000">自闭症谱系障碍Autism spectrum disorder</font>'''、'''<font color="#ff8000">精神分裂症Schizophrenia</font>'''、'''<font color="#ff8000">多动症ADHD<ref name=":5" /></font>'''和'''<font color="#ff8000">躁郁症Bipolar disorder<ref name=":6" /></font>'''。<br />
<br />
== Core networks==<br />
[[File:Heine2012x3010.png|thumb|An example that identified 10 large-scale brain networks from [[resting state fMRI]] activity through [[independent component analysis]].<ref name="Heine" />|链接=Special:FilePath/Heine2012x3010.png]]<br />
Because brain networks can be identified at various different resolutions and with various different neurobiological properties, there is no such thing as a universal atlas of brain networks that fits all circumstances.<ref name=":7">{{cite journal|last1=Eickhoff|first1=SB|last2=Yeo|first2=BTT|last3=Genon|first3=S|title=Imaging-based parcellations of the human brain.|journal=Nature Reviews. Neuroscience|date=November 2018|volume=19|issue=11|pages=672–686|doi=10.1038/s41583-018-0071-7|pmid=30305712|s2cid=52954265|url=http://juser.fz-juelich.de/record/856633/files/Eickhoff_Yeo_Genon_NRN_MainManuscriptInclFigures.pdf}}</ref> While acknowledging this problem, Uddin, Yeo, and Spreng proposed in 2019<ref name="Uddin2019">{{cite journal|last1=Uddin|first1=LQ|last2=Yeo|first2=BTT|last3=Spreng|first3=RN|title=Towards a Universal Taxonomy of Macro-scale Functional Human Brain Networks.|journal=Brain Topography|date=November 2019|volume=32|issue=6|pages=926–942|doi=10.1007/s10548-019-00744-6|pmid=31707621|pmc=7325607}}</ref> that the following six networks should be defined as core networks based on converging evidences from multiple studies<ref name=":8">{{cite journal|last1=Doucet|first1=GE|last2=Lee|first2=WH|last3=Frangou|first3=S|title=Evaluation of the spatial variability in the major resting-state networks across human brain functional atlases.|journal=Human Brain Mapping|date=2019-10-15|volume=40|issue=15|pages=4577–4587|doi=10.1002/hbm.24722|pmid=31322303|pmc=6771873}}</ref><ref name="Yeo" /><ref name=":9">{{cite journal|last1=Smith|first1=SM|last2=Fox|first2=PT|last3=Miller|first3=KL|last4=Glahn|first4=DC|last5=Fox|first5=PM|last6=Mackay|first6=CE|last7=Filippini|first7=N|last8=Watkins|first8=KE|last9=Toro|first9=R|last10=Laird|first10=AR|last11=Beckmann|first11=CF|title=Correspondence of the brain's functional architecture during activation and rest.|journal=Proceedings of the National Academy of Sciences of the United States of America|date=2009-08-04|volume=106|issue=31|pages=13040–5|doi=10.1073/pnas.0905267106|pmid=19620724|pmc=2722273|bibcode=2009PNAS..10613040S|doi-access=free}}</ref> to facilitate communication between researchers.<br />
<br />
【图1：An example that identified 10 large-scale brain networks from resting state fMRI activity through independent component analysis.<ref name="Heine" />这是一个通过独立元素分析方法从静息状态脑功能性磁共振成像信号中分辨出10个大规模脑网络的例子。<ref name="Heine" />】<br />
<br />
Because brain networks can be identified at various different resolutions and with various different neurobiological properties, there is no such thing as a universal atlas of brain networks that fits all circumstances. While acknowledging this problem, Uddin, Yeo, and Spreng proposed in 2019 that the following six networks should be defined as core networks based on converging evidences from multiple studies to facilitate communication between researchers.<br />
<br />
因为脑网络可以在不同分辨率条件下、使用不同的神经生物学特性来识别，所以没有能够适用于所有情况的通用脑网络图谱。<ref name=":7" />在承认这个问题的同时，Uddin，Yeo，和Spreng在2019年提出<ref name="Uddin2019" />，综合考虑来自多个研究的证据<ref name=":8" /><ref name="Yeo" /><ref name=":9" />，以下六个网络应该被定义为核心网络，以促进研究人员之间的交流。<br />
<br />
===Default Mode (Medial frontoparietal)===<br />
{{Main|Default mode network}}<br />
*The default mode network is active when an individual is awake and at rest. It preferentially activates when individuals focus on internally-oriented tasks such as daydreaming, envisioning the future, retrieving memories, and [[theory of mind]]. It is negatively correlated with brain systems that focus on external visual signals. It is the most widely researched network.<ref name="Bressler" /><ref name="Bassett" /><ref>{{Cite journal|date=2012-08-15|title=The serendipitous discovery of the brain's default network|journal=NeuroImage|language=en|volume=62|issue=2|pages=1137–1145|doi=10.1016/j.neuroimage.2011.10.035|pmid=22037421|issn=1053-8119|last1=Buckner|first1=Randy L.|s2cid=9880586}}</ref><ref name="Riedl" /><ref name="Yuan"><br />
{{cite journal|last1=Yuan|first1=Rui|last2=Di|first2=Xin|last3=Taylor|first3=Paul A.|last4=Gohel|first4=Suril|last5=Tsai|first5=Yuan-Hsiung|last6=Biswal|first6=Bharat B.|title=Functional topography of the thalamocortical system in human|journal=Brain Structure and Function|date=30 April 2015|doi=10.1007/s00429-015-1018-7|pmid=25924563|pmc=6363530|volume=221|issue=4|pages=1971–1984}}</ref><ref name="Bell">{{cite journal|last1=Bell|first1=Peter T.|last2=Shine|first2=James M.|title=Estimating Large-Scale Network Convergence in the Human Functional Connectome|journal=Brain Connectivity|date=2015-11-09|volume=5|issue=9|doi=10.1089/brain.2015.0348|pmid=26005099|pages=565–74}}</ref><ref name="Heine">{{cite journal|last1=Heine|first1=Lizette|last2=Soddu|first2=Andrea|last3=Gomez|first3=Francisco|last4=Vanhaudenhuyse|first4=Audrey|last5=Tshibanda|first5=Luaba|last6=Thonnard|first6=Marie|last7=Charland-Verville|first7=Vanessa|last8=Kirsch|first8=Murielle|last9=Laureys|first9=Steven|last10=Demertzi|first10=Athena|title=Resting state networks and consciousness. Alterations of multiple resting state network connectivity in physiological, pharmacological and pathological consciousness states.|journal=Frontiers in Psychology|date=2012|volume=3|pages=295|doi=10.3389/fpsyg.2012.00295|pmid=22969735|pmc=3427917|doi-access=free}}</ref><ref name="Yeo">{{cite journal|last1=Yeo|first1=B. T. Thomas|last2=Krienen|first2=Fenna M.|last3=Sepulcre|first3=Jorge|last4=Sabuncu|first4=Mert R.|last5=Lashkari|first5=Danial|last6=Hollinshead|first6=Marisa|last7=Roffman|first7=Joshua L.|last8=Smoller|first8=Jordan W.|last9=Zöllei|first9=Lilla|last10=Polimeni|first10=Jonathan R.|last11=Fischl|first11=Bruce|last12=Liu|first12=Hesheng|last13=Buckner|first13=Randy L.|title=The organization of the human cerebral cortex estimated by intrinsic functional connectivity|journal=Journal of Neurophysiology|date=2011-09-01|volume=106|issue=3|pages=1125–1165|doi=10.1152/jn.00338.2011|pmid=21653723|pmc=3174820|bibcode=2011NatSD...2E0031H }}</ref><ref name="Shafiei">{{cite journal|last1=Shafiei|first1=Golia|last2=Zeighami|first2=Yashar|last3=Clark|first3=Crystal A.|last4=Coull|first4=Jennifer T.|last5=Nagano-Saito|first5=Atsuko|last6=Leyton|first6=Marco|last7=Dagher|first7=Alain|last8=Mišić|first8=Bratislav|title=Dopamine Signaling Modulates the Stability and Integration of Intrinsic Brain Networks|journal=Cerebral Cortex|date=2018-10-01|volume=29|issue=1|pages=397–409|doi=10.1093/cercor/bhy264|pmid=30357316|pmc=6294404 }}</ref><ref name="Bailey">{{cite journal|last1=Bailey|first1=Stephen K.|last2=Aboud|first2=Katherine S.|last3=Nguyen|first3=Tin Q.|last4=Cutting|first4=Laurie E.|title=Applying a network framework to the neurobiology of reading and dyslexia|journal=Journal of Neurodevelopmental Disorders|date=13 December 2018|volume=10|issue=1|page=37|doi=10.1186/s11689-018-9251-z|pmid=30541433|pmc=6291929 }}</ref><br />
<br />
<br />
* The default mode network is active when an individual is awake and at rest. It preferentially activates when individuals focus on internally-oriented tasks such as daydreaming, envisioning the future, retrieving memories, and theory of mind. It is negatively correlated with brain systems that focus on external visual signals. It is the most widely researched network.<br />
<br />
<br />
=<nowiki>= = 默认模式网络（内侧额顶叶） = =</nowiki>= <br />
*默认模式网络在个体清醒和休息时都处于活跃状态。当个体专注于内部导向任务，比如做白日梦、展望未来、提取回忆和'''<font color="#ff8000">心智理论Theory of mind</font>'''任务时，默认模式网络会被优先激活。它与专注于外部视觉信号的大脑系统成负相关。对默认模式网络的研究目前是所谓网络之中最为广泛的。<br />
<br />
===Salience (Midcingulo-Insular)===<br />
{{Main|Salience network}}<br />
*The salience network consists of several structures, including the anterior (bilateral) insula, dorsal anterior cingulate cortex, and three subcortical structures which are the ventral striatum, substantia nigra/ventral tegmental region.<ref>{{Cite journal|last1=Steimke|first1=Rosa|last2=Nomi|first2=Jason S.|last3=Calhoun|first3=Vince D.|last4=Stelzel|first4=Christine|last5=Paschke|first5=Lena M.|last6=Gaschler|first6=Robert|last7=Goschke|first7=Thomas|last8=Walter|first8=Henrik|last9=Uddin|first9=Lucina Q.|date=2017-12-01|title=Salience network dynamics underlying successful resistance of temptation|journal=Social Cognitive and Affective Neuroscience|language=en|volume=12|issue=12|pages=1928–1939|doi=10.1093/scan/nsx123|pmid=29048582|pmc=5716209|issn=1749-5016|doi-access=free}}</ref><ref name=":0">{{Citation|last=Menon|first=V.|title=Brain Mapping|chapter=Salience Network|date=2015-01-01|chapter-url=http://www.sciencedirect.com/science/article/pii/B978012397025100052X|pages=597–611|editor-last=Toga|editor-first=Arthur W.|publisher=Academic Press|isbn=978-0-12-397316-0|access-date=2019-12-08|doi=10.1016/B978-0-12-397025-1.00052-X}}</ref> It plays the key role of monitoring the [[Salience (neuroscience)|salience]] of external inputs and internal brain events.<ref name="Riedl" /><ref name="Bressler" /><ref name="Bassett" /><ref name="Yuan" /><ref name="Heine" /><ref name="Yeo" /><ref name="Shafiei" /> Specifically, it aids in directing attention by identifying important biological and cognitive events.<ref name=":0" /><ref name="Bailey" /><br />
*This network includes the ventral attention network, which primarily includes the [[temporoparietal junction]] and the ventral [[frontal cortex]] of the right hemisphere.<ref name="Uddin2019" /><ref name="Vossel" /> These areas respond when behaviorally relevant stimuli occur unexpectedly.<ref name="Vossel" /> The ventral attention network is inhibited during focused attention in which top-down processing is being used, such as when visually searching for something. This response may prevent goal-driven attention from being distracted by non-relevant stimuli. It becomes active again when the target or relevant information about the target is found.<ref name="Vossel" /><ref>{{Cite journal|last1=Shulman|first1=Gordon L.|last2=McAvoy|first2=Mark P.|last3=Cowan|first3=Melanie C.|last4=Astafiev|first4=Serguei V.|last5=Tansy|first5=Aaron P.|last6=d'Avossa|first6=Giovanni|last7=Corbetta|first7=Maurizio|date=2003-11-01|title=Quantitative Analysis of Attention and Detection Signals During Visual Search|journal=Journal of Neurophysiology|volume=90|issue=5|pages=3384–3397|doi=10.1152/jn.00343.2003|pmid=12917383|issn=0022-3077}}</ref><br />
<br />
<br />
*The salience network consists of several structures, including the anterior (bilateral) insula, dorsal anterior cingulate cortex, and three subcortical structures which are the ventral striatum, substantia nigra/ventral tegmental region. It plays the key role of monitoring the salience of external inputs and internal brain events. Specifically, it aids in directing attention by identifying important biological and cognitive events.<br />
*This network includes the ventral attention network, which primarily includes the temporoparietal junction and the ventral frontal cortex of the right hemisphere. These areas respond when behaviorally relevant stimuli occur unexpectedly. The ventral attention network is inhibited during focused attention in which top-down processing is being used, such as when visually searching for something. This response may prevent goal-driven attention from being distracted by non-relevant stimuli. It becomes active again when the target or relevant information about the target is found.<br />
<br />
=<nowiki>= = 突显网络（扣带回-岛叶）= =</nowiki>= <br />
*突显网络由'''<font color="#32CD32"> 前(双)岛、前扣带回皮层和三个皮层下结构（腹侧纹状体、黑质/腹侧被盖区）</font>'''组成，它在监测外部输入和内部脑活动的'''<font color="#ff8000">凸显程度Salience</font>'''中发挥着关键作用。具体来说，突显网络通过识别重要的生理、认知活动来帮助引导注意力。<br />
*突显网络包括主要由右半球的'''<font color="#ff8000">颞顶联合区Temporoparietal junction</font>'''和腹侧'''<font color="#ff8000">额叶Frontal cortex</font>'''皮层组成的腹侧注意网络。当行为相关的刺激意外发生时，这些区域会对此做出反应。在自上而下加工注意焦点的过程中（例如使用视觉搜索某件物品时），腹侧注意网络会受到抑制。这种抑制反应可以防止目标驱动的注意力被不相关的刺激分散。当找到目标或相关信息时，突显网络会被再次激活。<br />
<br />
===Attention (Dorsal frontoparietal)===<br />
{{Main|Dorsal attention network}}<br />
*This network is involved in the voluntary, top-down deployment of attention.<ref name="Riedl" /><ref name="Yuan" /><ref name="Bell" /><ref name="Yeo" /><ref name="Shafiei" /><ref name="Vossel">{{cite journal|last1=Vossel|first1=Simone|last2=Geng|first2=Joy J.|last3=Fink|first3=Gereon R.|title=Dorsal and Ventral Attention Systems: Distinct Neural Circuits but Collaborative Roles|journal=The Neuroscientist|date=2014|volume=20|issue=2|pages=150–159|doi=10.1177/1073858413494269|pmid=23835449|pmc=4107817}}</ref><ref name="Hutton">{{cite journal|last1=Hutton|first1=John S.|last2=Dudley|first2=Jonathan|last3=Horowitz-Kraus|first3=Tzipi|last4=DeWitt|first4=Tom|last5=Holland|first5=Scott K.|title=Functional Connectivity of Attention, Visual, and Language Networks During Audio, Illustrated, and Animated Stories in Preschool-Age Children|journal=Brain Connectivity|date=1 September 2019|volume=9|issue=7|pages=580–592|doi=10.1089/brain.2019.0679|pmid=31144523|pmc=6775495|ref=Hutton}}</ref> Within the dorsal attention network, the intraparietal sulcus and frontal eye fields influence the visual areas of the brain. These influencing factors allow for the orientation of attention.<ref>{{Cite journal|last1=Fox|first1=Michael D.|last2=Corbetta|first2=Maurizio|last3=Snyder|first3=Abraham Z.|last4=Vincent|first4=Justin L.|last5=Raichle|first5=Marcus E.|date=2006-06-27|title=Spontaneous neuronal activity distinguishes human dorsal and ventral attention systems|journal=Proceedings of the National Academy of Sciences|language=en|volume=103|issue=26|pages=10046–10051|doi=10.1073/pnas.0604187103|issn=0027-8424|pmid=16788060|pmc=1480402|bibcode=2006PNAS..10310046F|doi-access=free}}</ref><ref name="Vossel" /><ref name="Bailey" /><br />
<br />
<br />
*This network is involved in the voluntary, top-down deployment of attention. Within the dorsal attention network, the intraparietal sulcus and frontal eye fields influence the visual areas of the brain. These influencing factors allow for the orientation of attention.<br />
<br />
=<nowiki>= = 注意网络(背侧额顶骨) = =</nowiki>= <br />
*这个网络参与了自发的、自上而下的注意力分配。在背侧注意网络中，顶内沟和额眼影响着大脑的视觉区域。这些影响因素决定了注意力的方向。<br />
<br />
===Control (Lateral frontoparietal)===<br />
{{Main|Frontoparietal network}}<br />
*This network initiates and modulates cognitive control and comprises 18 sub-regions of the brain.<ref>{{Cite journal|last1=Scolari|first1=Miranda|last2=Seidl-Rathkopf|first2=Katharina N|last3=Kastner|first3=Sabine|date=2015-02-01|title=Functions of the human frontoparietal attention network: Evidence from neuroimaging|journal=Current Opinion in Behavioral Sciences|series=Cognitive control|volume=1|pages=32–39|doi=10.1016/j.cobeha.2014.08.003|issn=2352-1546|pmid=27398396|pmc=4936532}}</ref> There is a strong correlation between fluid intelligence and the involvement of the fronto-parietal network with other networks.<ref>{{Cite journal|last1=Marek|first1=Scott|last2=Dosenbach|first2=Nico U. F.|date=June 2018|title=The frontoparietal network: function, electrophysiology, and importance of individual precision mapping|journal=Dialogues in Clinical Neuroscience|volume=20|issue=2|pages=133–140|doi=10.31887/DCNS.2018.20.2/smarek|issn=1294-8322|pmc=6136121|pmid=30250390}}</ref><br />
<br />
*Versions of this network have also been called the central executive (or executive control) network and the cognitive control network.<ref name="Uddin2019" /><br />
<br />
<br />
*This network initiates and modulates cognitive control and comprises 18 sub-regions of the brain. There is a strong correlation between fluid intelligence and the involvement of the fronto-parietal network with other networks.<br />
<br />
=<nowiki>= = 控制网络(侧额顶骨) = =</nowiki>= <br />
*这个网络参与了认知控制的启动与调节，它包括了大脑的18个亚区。额顶网络与其他网络的参与程度和流体智力之间存在着有很强的相关性。<br />
<br />
<br />
*Versions of this network have also been called the central executive (or executive control) network and the cognitive control network.<br />
<br />
*在其它版本中，这种网络也被称为中央执行(或执行控制)网络和认知控制网络。<br />
<br />
===Sensorimotor or Somatomotor (Pericentral)===<br />
{{Main|Sensorimotor network}}<br />
*This network processes somatosensory information and coordinates motion.<ref name="Heine" /><ref name="Yeo" /><ref name="Shafiei" /><ref name="Bassett" /><ref name="Yuan" /> The [[auditory cortex]] may be included.<ref name="Uddin2019" /><ref name="Yeo" /><br />
<br />
*This network processes somatosensory information and coordinates motion. The auditory cortex may be included.<br />
<br />
=<nowiki>= = 感觉运动网络(中央区域) = =</nowiki>= <br />
* 这个网络参与了躯体感觉信息的加工和运动的协调。听觉皮层可能也包括在内。<br />
<br />
===Visual (Occipital) ===<br />
{{See|Visual cortex}}<br />
*This network handles visual information processing.<ref name="Yang">{{cite journal|last1=Yang|first1=Yan-li|last2=Deng|first2=Hong-xia|last3=Xing|first3=Gui-yang|last4=Xia|first4=Xiao-luan|last5=Li|first5=Hai-fang|title=Brain functional network connectivity based on a visual task: visual information processing-related brain regions are significantly activated in the task state|journal=Neural Regeneration Research|date=2015|volume=10|issue=2|pages=298–307|doi=10.4103/1673-5374.152386|pmid=25883631|pmc=4392680 }}</ref><br />
<br />
<br />
*This network handles visual information processing.<br />
<br />
=<nowiki>= = 视觉网络(枕部) = =</nowiki>= <br />
*这个网络处理视觉信息。<br />
<br />
== Other networks==<br />
Different methods and data have identified several other brain networks, many of which greatly overlap or are subsets of more well-characterized core networks.<ref name="Uddin2019" /><br />
*Limbic<ref name="Bassett" /><ref name="Yeo" /><ref name="Bailey" /><br />
*Auditory<ref name="Yuan" /><ref name="Heine" /><br />
*Right/left executive<ref name="Yuan" /><ref name="Heine" /><br />
*Cerebellar<ref name="Bell" /><ref name="Heine" /><br />
*Spatial attention<ref name="Riedl" /><ref name="Bressler" /><br />
*Language<ref name="Bressler" /><ref name="Hutton" /><br />
* Lateral visual<ref name="Yuan" /><ref name="Bell" /><ref name="Heine" /><br />
*Temporal<ref name="Yeo" /><ref name="Shafiei" /><br />
*Visual perception/imagery<ref name="Hutton" /><br />
<br />
Different methods and data have identified several other brain networks, many of which greatly overlap or are subsets of more well-characterized core networks.<br />
*Limbic<br />
*Auditory<br />
*Right/left executive<br />
*Cerebellar<br />
*Spatial attention<br />
*Language<br />
*Lateral visual<br />
*Temporal<br />
*Visual perception/imagery<br />
<br />
其他网络不同的方法和数据已经确定了其他几个大脑网络，其中许多网络极大地重叠或者是更具特色的核心网络的子集。<br />
*边缘<br />
*听觉<br />
*右/左执行<br />
*小脑<br />
*空间注意<br />
*语言<br />
*外侧视觉<br />
*颞视知觉/图像<br />
<br />
==See also==<br />
*[[Complex network]]<br />
*[[Neural network]]<br />
<br />
*Complex network<br />
*Neural network<br />
<br />
=<nowiki>= = =</nowiki>= <br />
*复杂网络<br />
*<br />
*神经网络<br />
*<br />
<br />
==References==<br />
{{reflist|30em}}<br />
<br />
{{Human connectomics}}<br />
<br />
[[Category:Neuroscience]]<br />
[[Category:Neural coding]]<br />
[[Category:Neural circuits]]<br />
[[Category:Neurophysiology]]<br />
<br />
Category:Neuroscience<br />
Category:Neural coding<br />
Category:Neural circuits<br />
Category:Neurophysiology<br />
<br />
类别: 神经科学类别: 神经编码类别: 神经回路类别: 神经生理学<br />
<br />
<noinclude><br />
<br />
<small>This page was moved from [[wikipedia:en:Large-scale brain network]]. Its edit history can be viewed at [[大规模脑网络/edithistory]]</small></noinclude><br />
<br />
[[Category:待整理页面]]</div>Shenky20https://wiki.swarma.org/index.php?title=%E5%A4%A7%E8%A7%84%E6%A8%A1%E8%84%91%E7%BD%91%E7%BB%9C&diff=29989大规模脑网络2022-04-04T08:56:50Z<p>Shenky20：</p>
<hr />
<div>此词条由神经动力学模型读书会词条梳理志愿者Shenky20翻译审校，翻译字数共800，未经专家审核，带来阅读不便，请见谅。<br />
<br />
{{Short description|Collections of brain regions}}<br />
'''Large-scale brain networks''' (also known as '''intrinsic brain networks''') are collections of widespread [[brain regions]] showing [[Resting state fMRI#Functional|functional connectivity]] by statistical analysis of the [[Functional magnetic resonance imaging|fMRI]] [[BOLD signal]]<ref name=Riedl>{{cite journal|last1=Riedl|first1=Valentin|last2=Utz|first2=Lukas|last3=Castrillón|first3=Gabriel|last4=Grimmer|first4=Timo|last5=Rauschecker|first5=Josef P.|last6=Ploner|first6=Markus|last7=Friston|first7=Karl J.|last8=Drzezga|first8=Alexander|last9=Sorg|first9=Christian|title=Metabolic connectivity mapping reveals effective connectivity in the resting human brain|journal=PNAS|date=January 12, 2016|volume=113|issue=2|pages=428–433|doi=10.1073/pnas.1513752113|pmid=26712010|pmc=4720331|bibcode=2016PNAS..113..428R|doi-access=free}}</ref> or other recording methods such as [[Electroencephalography|EEG]],<ref name=":1">{{Cite journal|last1=Foster|first1=Brett L.|last2=Parvizi|first2=Josef|date=2012-03-01|title=Resting oscillations and cross-frequency coupling in the human posteromedial cortex|journal=NeuroImage|volume=60|issue=1|pages=384–391|doi=10.1016/j.neuroimage.2011.12.019|pmid=22227048|issn=1053-8119|pmc=3596417}}</ref> [[Positron emission tomography|PET]]<ref name=":2">{{Cite journal|last1=Buckner|first1=Randy L.|last2=Andrews‐Hanna|first2=Jessica R.|last3=Schacter|first3=Daniel L.|date=2008|title=The Brain's Default Network|journal=Annals of the New York Academy of Sciences|language=en|volume=1124|issue=1|pages=1–38|doi=10.1196/annals.1440.011|pmid=18400922|issn=1749-6632|bibcode=2008NYASA1124....1B|s2cid=3167595}}</ref> and [[Magnetoencephalography|MEG]].<ref name=":3">{{Cite journal|last1=Morris|first1=Peter G.|last2=Smith|first2=Stephen M.|last3=Barnes|first3=Gareth R.|last4=Stephenson|first4=Mary C.|last5=Hale|first5=Joanne R.|last6=Price|first6=Darren|last7=Luckhoo|first7=Henry|last8=Woolrich|first8=Mark|last9=Brookes|first9=Matthew J.|date=2011-10-04|title=Investigating the electrophysiological basis of resting state networks using magnetoencephalography|journal=Proceedings of the National Academy of Sciences|language=en|volume=108|issue=40|pages=16783–16788|doi=10.1073/pnas.1112685108|issn=0027-8424|pmid=21930901|pmc=3189080|bibcode=2011PNAS..10816783B|doi-access=free}}</ref> An emerging paradigm in neuroscience is that cognitive tasks are performed not by individual brain regions working in isolation but by networks consisting of several discrete brain regions that are said to be "functionally connected". Functional connectivity networks may be found using algorithms such as [[cluster analysis]], spatial [[independent component analysis]] (ICA), seed based, and others.<ref name="Petersen">{{cite journal|last1=Petersen|first1=Steven|last2=Sporns|first2=Olaf|title=Brain Networks and Cognitive Architectures|journal=Neuron|date=October 2015|volume=88|issue=1|pages=207–219|doi=10.1016/j.neuron.2015.09.027|pmid=26447582|pmc=4598639 }}</ref> Synchronized brain regions may also be identified using long-range synchronization of the EEG, MEG, or other dynamic brain signals.<ref name=Bressler>{{cite journal|last1=Bressler|first1=Steven L.|last2=Menon|first2=Vinod|s2cid=5967761|title=Large scale brain networks in cognition: emerging methods and principles|journal=Trends in Cognitive Sciences|date=June 2010|volume=14|issue=6|pages=233–290|doi=10.1016/j.tics.2010.04.004|url=http://www.cell.com/trends/cognitive-sciences/issue?pii=S1364-6613(10)X0005-5|accessdate=24 January 2016|pmid=20493761}}</ref><br />
<br />
<br />
Large-scale brain networks (also known as intrinsic brain networks) are collections of widespread brain regions showing functional connectivity by statistical analysis of the fMRI BOLD signal or other recording methods such as EEG, PET and MEG. An emerging paradigm in neuroscience is that cognitive tasks are performed not by individual brain regions working in isolation but by networks consisting of several discrete brain regions that are said to be "functionally connected". Functional connectivity networks may be found using algorithms such as cluster analysis, spatial independent component analysis (ICA), seed based, and others. Synchronized brain regions may also be identified using long-range synchronization of the EEG, MEG, or other dynamic brain signals.<br />
<br />
'''<font color="#ff8000">大规模脑网络Large-scale brain networks</font>'''(也称为'''<font color="#ff8000">内在大脑网络Intrinsic brain networks</font>''')是在对基于'''<font color="#ff8000">血氧水平依赖效应BOLD</font>'''的'''<font color="#ff8000">功能性磁共振成像fMRI</font>'''信号<ref name="Riedl" />的统计分析或其他记录方法(如'''<font color="#ff8000">脑电图EEG<ref name=":1" /></font>'''、'''<font color="#ff8000">正电子发射断层扫描技术PET<ref name=":2" /></font>'''和'''<font color="#ff8000">脑磁图MEG<ref name=":3" /></font>''')中，表现出'''<font color="#ff8000">功能连接Functional connectivity</font>'''的'''<font color="#ff8000">脑区Brain regions</font>'''的集合。根据神经科学中一个新出现的范式，认知任务不是由单个脑区独立执行的，而是由几个互不相连的脑区“功能连接”组成的网络执行的。功能连接网络可以通过'''<font color="#ff8000">数据聚类Cluster analysis</font>'''、空间'''<font color="#ff8000">独立元素分析ICA</font>'''、种子点方法等算法来发现。<ref name="Petersen" />同步的脑区也可以用脑电图、脑磁图或其他动态脑信号的远程同步来识别。<ref name="Bressler" /><br />
<br />
The set of identified brain areas that are linked together in a large-scale network varies with cognitive function.<ref name="Bressler2">{{cite journal|last1=Bressler|first1=Steven L.|title=Neurocognitive networks|journal=Scholarpedia|volume=3|issue=2|pages=1567|doi=10.4249/scholarpedia.1567|year=2008|bibcode=2008SchpJ...3.1567B|doi-access=free}}</ref> When the cognitive state is not explicit (i.e., the subject is at "rest"), the large-scale brain network is a [[Resting state fMRI|resting state]] network (RSN). As a physical system with graph-like properties,<ref name="Bressler" /> a large-scale brain network has both nodes and edges and cannot be identified simply by the co-activation of brain areas. In recent decades, the analysis of brain networks was made feasible by advances in imaging techniques as well as new tools from [[graph theory]] and [[Dynamical systems theory|dynamical systems]].<br />
<br />
The set of identified brain areas that are linked together in a large-scale network varies with cognitive function. When the cognitive state is not explicit (i.e., the subject is at "rest"), the large-scale brain network is a resting state network (RSN). As a physical system with graph-like properties, a large-scale brain network has both nodes and edges and cannot be identified simply by the co-activation of brain areas. In recent decades, the analysis of brain networks was made feasible by advances in imaging techniques as well as new tools from graph theory and dynamical systems.<br />
<br />
大规模脑网络中,连接在一起的脑区集合因认知功能的不同而不同。<ref name="Bressler2" />当认知状态不明确(即主体处于“静止”状态)时，大规模脑网络是一个'''<font color="#ff8000">静息状态Resting state</font>'''网络(RSN)。作为一个具有图形特征的物理系统，<ref name="Bressler" />大规模脑网络既有节点又有边，不能简单地通过脑区的共同激活来识别。近几十年来，成像技术不断进步，同时'''<font color="#ff8000">图论Graph theory</font>'''、'''<font color="#ff8000">动力学系统Dynamical systems</font>'''领域出现了新的技术手段，这使得脑网络分析变得可行。<br />
<br />
Large-scale brain networks are identified by their function and provide a coherent framework for understanding [[cognition]] by offering a neural model of how different cognitive functions emerge when different sets of brain regions join together as self-organized coalitions. The number and composition of the coalitions will vary with the algorithm and parameters used to identify them.<ref name="Yeo" /><ref name=":4">{{cite journal|last1=Abou Elseoud|first1=Ahmed|last2=Littow|first2=Harri|last3=Remes|first3=Jukka|last4=Starck|first4=Tuomo|last5=Nikkinen|first5=Juha|last6=Nissilä|first6=Juuso|last7=Timonen|first7=Markku|last8=Tervonen|first8=Osmo|last9=Kiviniemi1|first9=Vesa|title=Group-ICA Model Order Highlights Patterns of Functional Brain Connectivity|journal= Frontiers in Systems Neuroscience|date=2011-06-03|volume=5|pages=37|doi=10.3389/fnsys.2011.00037|pmid=21687724|pmc=3109774 |doi-access=free}}</ref> In one model, there is only the [[default mode network]] and the [[task-positive network]], but most current analyses show several networks, from a small handful to 17.<ref name="Yeo" /> The most common and stable networks are enumerated below. The regions participating in a functional network may be dynamically reconfigured.<ref name="Petersen" /><ref name="Bassett">{{cite journal|last1=Bassett|first1=Daniella|last2=Bertolero|first2=Max|title=How Matter Becomes Mind|journal=Scientific American|date=July 2019|volume=321|issue=1|page=32|url=https://www.scientificamerican.com/|accessdate=23 June 2019 }}</ref><br />
<br />
Large-scale brain networks are identified by their function and provide a coherent framework for understanding cognition by offering a neural model of how different cognitive functions emerge when different sets of brain regions join together as self-organized coalitions. The number and composition of the coalitions will vary with the algorithm and parameters used to identify them. In one model, there is only the default mode network and the task-positive network, but most current analyses show several networks, from a small handful to 17. The most common and stable networks are enumerated below. The regions participating in a functional network may be dynamically reconfigured.<br />
<br />
大规模脑网络是通过其功能来进行识别的。通过研究大规模脑网络建立神经模型，对不同脑区组合所形成的自组织联合体如何实现不同的'''<font color="#ff8000">认知Cognition</font>'''功能进行解释，就能够为认知理解提供一个连贯的框架。识别算法和参数的不同会导致所识别出的上述联合体的数量和组成有所不同。<ref name="Yeo" /><ref name=":4" />一个模型理论认为，符合上述条件的神经模型只包含'''<font color="#ff8000">默认模式网络Default mode network</font>'''和'''<font color="#ff8000">任务激活网络Task-positive network</font>'''，但目前'''<font color="#32CD32">大多数分析理论都包括从几个到17个不等的网络。<ref name="Yeo" /></font>'''下面列举了最常见且稳定的网络。'''<font color="#32CD32"> 人脑</font>'''可以动态地重新配置参与功能网络的脑区。<ref name="Petersen" /><ref name="Bassett" /> <br />
<br />
Disruptions in activity in various networks have been implicated in neuropsychiatric disorders such as [[Depression (mood)|depression]], [[Alzheimer's disease|Alzheimer's]], [[Autism-spectrum disorder|autism spectrum disorder]], [[schizophrenia]], [[ADHD]]<ref name=":5">{{cite journal |last1=Griffiths |first1=Kristi R. |last2=Braund |first2=Taylor A. |last3=Kohn |first3=Michael R. |last4=Clarke |first4=Simon |last5=Williams |first5=Leanne M. |last6=Korgaonkar |first6=Mayuresh S. |title=Structural brain network topology underpinning ADHD and response to methylphenidate treatment |journal=Translational Psychiatry |date=2 March 2021 |volume=11 |issue=1 |pages=1–9 |doi=10.1038/s41398-021-01278-x | pmc=7925571 |pmid=33654073 |url=https://www.nature.com/articles/s41398-021-01278-x#citeas |access-date=16 November 2021}}</ref> and [[bipolar disorder]].<ref name=":6">{{Cite journal|url=https://www.researchgate.net/publication/51639686|title=Large-scale brain networks and psychopathology: A unifying triple network model|last=Menon|first=Vinod|s2cid=26653572|journal=Trends in Cognitive Sciences|date=2011-09-09|volume=15|issue=10|pages=483–506|doi=10.1016/j.tics.2011.08.003|pmid=21908230}}</ref><br />
<br />
Disruptions in activity in various networks have been implicated in neuropsychiatric disorders such as depression, Alzheimer's, autism spectrum disorder, schizophrenia, ADHD and bipolar disorder.<br />
<br />
脑网络活动的中断与诸多神经精神疾病密切相关，如'''<font color="#ff8000">抑郁症Depression</font>'''、'''<font color="#ff8000">老年痴呆症Alzheimer's</font>'''、'''<font color="#ff8000">自闭症谱系障碍Autism spectrum disorder</font>'''、'''<font color="#ff8000">精神分裂症Schizophrenia</font>'''、'''<font color="#ff8000">多动症ADHD<ref name=":5" /></font>'''和'''<font color="#ff8000">躁郁症Bipolar disorder<ref name=":6" /></font>'''。<br />
<br />
== Core networks==<br />
[[File:Heine2012x3010.png|thumb|An example that identified 10 large-scale brain networks from [[resting state fMRI]] activity through [[independent component analysis]].<ref name="Heine" />|链接=Special:FilePath/Heine2012x3010.png]]<br />
Because brain networks can be identified at various different resolutions and with various different neurobiological properties, there is no such thing as a universal atlas of brain networks that fits all circumstances.<ref name=":7">{{cite journal|last1=Eickhoff|first1=SB|last2=Yeo|first2=BTT|last3=Genon|first3=S|title=Imaging-based parcellations of the human brain.|journal=Nature Reviews. Neuroscience|date=November 2018|volume=19|issue=11|pages=672–686|doi=10.1038/s41583-018-0071-7|pmid=30305712|s2cid=52954265|url=http://juser.fz-juelich.de/record/856633/files/Eickhoff_Yeo_Genon_NRN_MainManuscriptInclFigures.pdf}}</ref> While acknowledging this problem, Uddin, Yeo, and Spreng proposed in 2019<ref name="Uddin2019">{{cite journal|last1=Uddin|first1=LQ|last2=Yeo|first2=BTT|last3=Spreng|first3=RN|title=Towards a Universal Taxonomy of Macro-scale Functional Human Brain Networks.|journal=Brain Topography|date=November 2019|volume=32|issue=6|pages=926–942|doi=10.1007/s10548-019-00744-6|pmid=31707621|pmc=7325607}}</ref> that the following six networks should be defined as core networks based on converging evidences from multiple studies<ref name=":8">{{cite journal|last1=Doucet|first1=GE|last2=Lee|first2=WH|last3=Frangou|first3=S|title=Evaluation of the spatial variability in the major resting-state networks across human brain functional atlases.|journal=Human Brain Mapping|date=2019-10-15|volume=40|issue=15|pages=4577–4587|doi=10.1002/hbm.24722|pmid=31322303|pmc=6771873}}</ref><ref name="Yeo" /><ref name=":9">{{cite journal|last1=Smith|first1=SM|last2=Fox|first2=PT|last3=Miller|first3=KL|last4=Glahn|first4=DC|last5=Fox|first5=PM|last6=Mackay|first6=CE|last7=Filippini|first7=N|last8=Watkins|first8=KE|last9=Toro|first9=R|last10=Laird|first10=AR|last11=Beckmann|first11=CF|title=Correspondence of the brain's functional architecture during activation and rest.|journal=Proceedings of the National Academy of Sciences of the United States of America|date=2009-08-04|volume=106|issue=31|pages=13040–5|doi=10.1073/pnas.0905267106|pmid=19620724|pmc=2722273|bibcode=2009PNAS..10613040S|doi-access=free}}</ref> to facilitate communication between researchers.<br />
<br />
【图1：An example that identified 10 large-scale brain networks from resting state fMRI activity through independent component analysis.<ref name="Heine" />这是一个通过独立元素分析方法从静息状态脑功能性磁共振成像信号中分辨出10个大规模脑网络的例子。<ref name="Heine" />】<br />
<br />
Because brain networks can be identified at various different resolutions and with various different neurobiological properties, there is no such thing as a universal atlas of brain networks that fits all circumstances. While acknowledging this problem, Uddin, Yeo, and Spreng proposed in 2019 that the following six networks should be defined as core networks based on converging evidences from multiple studies to facilitate communication between researchers.<br />
<br />
因为脑网络可以在不同分辨率条件下、使用不同的神经生物学特性来识别，所以没有能够适用于所有情况的通用脑网络图谱。<ref name=":7" />在承认这个问题的同时，Uddin，Yeo，和Spreng在2019年提出<ref name="Uddin2019" />，综合考虑来自多个研究的证据<ref name=":8" /><ref name="Yeo" /><ref name=":9" />，以下六个网络应该被定义为核心网络，以促进研究人员之间的交流。<br />
<br />
===Default Mode (Medial frontoparietal)===<br />
{{Main|Default mode network}}<br />
*The default mode network is active when an individual is awake and at rest. It preferentially activates when individuals focus on internally-oriented tasks such as daydreaming, envisioning the future, retrieving memories, and [[theory of mind]]. It is negatively correlated with brain systems that focus on external visual signals. It is the most widely researched network.<ref name="Bressler" /><ref name="Bassett" /><ref>{{Cite journal|date=2012-08-15|title=The serendipitous discovery of the brain's default network|journal=NeuroImage|language=en|volume=62|issue=2|pages=1137–1145|doi=10.1016/j.neuroimage.2011.10.035|pmid=22037421|issn=1053-8119|last1=Buckner|first1=Randy L.|s2cid=9880586}}</ref><ref name="Riedl" /><ref name="Yuan"><br />
{{cite journal|last1=Yuan|first1=Rui|last2=Di|first2=Xin|last3=Taylor|first3=Paul A.|last4=Gohel|first4=Suril|last5=Tsai|first5=Yuan-Hsiung|last6=Biswal|first6=Bharat B.|title=Functional topography of the thalamocortical system in human|journal=Brain Structure and Function|date=30 April 2015|doi=10.1007/s00429-015-1018-7|pmid=25924563|pmc=6363530|volume=221|issue=4|pages=1971–1984}}</ref><ref name="Bell">{{cite journal|last1=Bell|first1=Peter T.|last2=Shine|first2=James M.|title=Estimating Large-Scale Network Convergence in the Human Functional Connectome|journal=Brain Connectivity|date=2015-11-09|volume=5|issue=9|doi=10.1089/brain.2015.0348|pmid=26005099|pages=565–74}}</ref><ref name="Heine">{{cite journal|last1=Heine|first1=Lizette|last2=Soddu|first2=Andrea|last3=Gomez|first3=Francisco|last4=Vanhaudenhuyse|first4=Audrey|last5=Tshibanda|first5=Luaba|last6=Thonnard|first6=Marie|last7=Charland-Verville|first7=Vanessa|last8=Kirsch|first8=Murielle|last9=Laureys|first9=Steven|last10=Demertzi|first10=Athena|title=Resting state networks and consciousness. Alterations of multiple resting state network connectivity in physiological, pharmacological and pathological consciousness states.|journal=Frontiers in Psychology|date=2012|volume=3|pages=295|doi=10.3389/fpsyg.2012.00295|pmid=22969735|pmc=3427917|doi-access=free}}</ref><ref name="Yeo">{{cite journal|last1=Yeo|first1=B. T. Thomas|last2=Krienen|first2=Fenna M.|last3=Sepulcre|first3=Jorge|last4=Sabuncu|first4=Mert R.|last5=Lashkari|first5=Danial|last6=Hollinshead|first6=Marisa|last7=Roffman|first7=Joshua L.|last8=Smoller|first8=Jordan W.|last9=Zöllei|first9=Lilla|last10=Polimeni|first10=Jonathan R.|last11=Fischl|first11=Bruce|last12=Liu|first12=Hesheng|last13=Buckner|first13=Randy L.|title=The organization of the human cerebral cortex estimated by intrinsic functional connectivity|journal=Journal of Neurophysiology|date=2011-09-01|volume=106|issue=3|pages=1125–1165|doi=10.1152/jn.00338.2011|pmid=21653723|pmc=3174820|bibcode=2011NatSD...2E0031H }}</ref><ref name="Shafiei">{{cite journal|last1=Shafiei|first1=Golia|last2=Zeighami|first2=Yashar|last3=Clark|first3=Crystal A.|last4=Coull|first4=Jennifer T.|last5=Nagano-Saito|first5=Atsuko|last6=Leyton|first6=Marco|last7=Dagher|first7=Alain|last8=Mišić|first8=Bratislav|title=Dopamine Signaling Modulates the Stability and Integration of Intrinsic Brain Networks|journal=Cerebral Cortex|date=2018-10-01|volume=29|issue=1|pages=397–409|doi=10.1093/cercor/bhy264|pmid=30357316|pmc=6294404 }}</ref><ref name="Bailey">{{cite journal|last1=Bailey|first1=Stephen K.|last2=Aboud|first2=Katherine S.|last3=Nguyen|first3=Tin Q.|last4=Cutting|first4=Laurie E.|title=Applying a network framework to the neurobiology of reading and dyslexia|journal=Journal of Neurodevelopmental Disorders|date=13 December 2018|volume=10|issue=1|page=37|doi=10.1186/s11689-018-9251-z|pmid=30541433|pmc=6291929 }}</ref><br />
<br />
<br />
* The default mode network is active when an individual is awake and at rest. It preferentially activates when individuals focus on internally-oriented tasks such as daydreaming, envisioning the future, retrieving memories, and theory of mind. It is negatively correlated with brain systems that focus on external visual signals. It is the most widely researched network.<br />
<br />
<br />
=<nowiki>= = 默认模式网络（'''<font color="#32CD32"> 内侧额顶叶</font>''' ） = =</nowiki>= <br />
*默认模式网络在个体清醒和休息时都处于活跃状态。当个体专注于内部导向任务，比如做白日梦、展望未来、提取回忆和'''<font color="#ff8000">心智理论Theory of mind</font>'''任务时，默认模式网络会被优先激活。它与专注于外部视觉信号的大脑系统成负相关。对默认模式网络的研究目前是所谓网络之中最为广泛的。<br />
<br />
=== Salience (Midcingulo-Insular)===<br />
{{Main|Salience network}}<br />
* The salience network consists of several structures, including the anterior (bilateral) insula, dorsal anterior cingulate cortex, and three subcortical structures which are the ventral striatum, substantia nigra/ventral tegmental region.<ref>{{Cite journal|last1=Steimke|first1=Rosa|last2=Nomi|first2=Jason S.|last3=Calhoun|first3=Vince D.|last4=Stelzel|first4=Christine|last5=Paschke|first5=Lena M.|last6=Gaschler|first6=Robert|last7=Goschke|first7=Thomas|last8=Walter|first8=Henrik|last9=Uddin|first9=Lucina Q.|date=2017-12-01|title=Salience network dynamics underlying successful resistance of temptation|journal=Social Cognitive and Affective Neuroscience|language=en|volume=12|issue=12|pages=1928–1939|doi=10.1093/scan/nsx123|pmid=29048582|pmc=5716209|issn=1749-5016|doi-access=free}}</ref><ref name=":0">{{Citation|last=Menon|first=V.|title=Brain Mapping|chapter=Salience Network|date=2015-01-01|chapter-url=http://www.sciencedirect.com/science/article/pii/B978012397025100052X|pages=597–611|editor-last=Toga|editor-first=Arthur W.|publisher=Academic Press|isbn=978-0-12-397316-0|access-date=2019-12-08|doi=10.1016/B978-0-12-397025-1.00052-X}}</ref> It plays the key role of monitoring the [[Salience (neuroscience)|salience]] of external inputs and internal brain events.<ref name="Riedl" /><ref name="Bressler" /><ref name="Bassett" /><ref name="Yuan" /><ref name="Heine" /><ref name="Yeo" /><ref name="Shafiei" /> Specifically, it aids in directing attention by identifying important biological and cognitive events.<ref name=":0" /><ref name="Bailey" /><br />
*This network includes the ventral attention network, which primarily includes the [[temporoparietal junction]] and the ventral [[frontal cortex]] of the right hemisphere.<ref name="Uddin2019" /><ref name="Vossel" /> These areas respond when behaviorally relevant stimuli occur unexpectedly.<ref name="Vossel" /> The ventral attention network is inhibited during focused attention in which top-down processing is being used, such as when visually searching for something. This response may prevent goal-driven attention from being distracted by non-relevant stimuli. It becomes active again when the target or relevant information about the target is found.<ref name="Vossel" /><ref>{{Cite journal|last1=Shulman|first1=Gordon L.|last2=McAvoy|first2=Mark P.|last3=Cowan|first3=Melanie C.|last4=Astafiev|first4=Serguei V.|last5=Tansy|first5=Aaron P.|last6=d'Avossa|first6=Giovanni|last7=Corbetta|first7=Maurizio|date=2003-11-01|title=Quantitative Analysis of Attention and Detection Signals During Visual Search|journal=Journal of Neurophysiology|volume=90|issue=5|pages=3384–3397|doi=10.1152/jn.00343.2003|pmid=12917383|issn=0022-3077}}</ref><br />
<br />
<br />
* The salience network consists of several structures, including the anterior (bilateral) insula, dorsal anterior cingulate cortex, and three subcortical structures which are the ventral striatum, substantia nigra/ventral tegmental region. It plays the key role of monitoring the salience of external inputs and internal brain events. Specifically, it aids in directing attention by identifying important biological and cognitive events.<br />
* This network includes the ventral attention network, which primarily includes the temporoparietal junction and the ventral frontal cortex of the right hemisphere. These areas respond when behaviorally relevant stimuli occur unexpectedly. The ventral attention network is inhibited during focused attention in which top-down processing is being used, such as when visually searching for something. This response may prevent goal-driven attention from being distracted by non-relevant stimuli. It becomes active again when the target or relevant information about the target is found.<br />
<br />
=<nowiki>= = 突显网络（'''<font color="#32CD32"> 扣带回-岛叶</font>'''）= =</nowiki>= <br />
*突显网络由'''<font color="#32CD32"> 前(双)岛、前扣带回皮层和三个皮层下结构（腹侧纹状体、黑质/腹侧被盖区）</font>'''组成，它在监测外部输入和内部脑活动的'''<font color="#ff8000">凸显程度Salience</font>'''中发挥着关键作用。具体来说，突显网络通过识别重要的生理、认知活动来帮助引导注意力。<br />
*突显网络包括主要由右半球的'''<font color="#ff8000">颞顶联合区Temporoparietal junction</font>'''和腹侧'''<font color="#ff8000">额叶Frontal cortex</font>'''皮层组成的腹侧注意网络。当行为相关的刺激意外发生时，这些区域会对此做出反应。在自上而下加工注意焦点的过程中（例如使用视觉搜索某件物品时），腹侧注意网络会受到抑制。这种抑制反应可以防止目标驱动的注意力被不相关的刺激分散。当找到目标或相关信息时，突显网络会被再次激活。<br />
<br />
===Attention (Dorsal frontoparietal)===<br />
{{Main|Dorsal attention network}}<br />
*This network is involved in the voluntary, top-down deployment of attention.<ref name="Riedl" /><ref name="Yuan" /><ref name="Bell" /><ref name="Yeo" /><ref name="Shafiei" /><ref name="Vossel">{{cite journal|last1=Vossel|first1=Simone|last2=Geng|first2=Joy J.|last3=Fink|first3=Gereon R.|title=Dorsal and Ventral Attention Systems: Distinct Neural Circuits but Collaborative Roles|journal=The Neuroscientist|date=2014|volume=20|issue=2|pages=150–159|doi=10.1177/1073858413494269|pmid=23835449|pmc=4107817}}</ref><ref name="Hutton">{{cite journal|last1=Hutton|first1=John S.|last2=Dudley|first2=Jonathan|last3=Horowitz-Kraus|first3=Tzipi|last4=DeWitt|first4=Tom|last5=Holland|first5=Scott K.|title=Functional Connectivity of Attention, Visual, and Language Networks During Audio, Illustrated, and Animated Stories in Preschool-Age Children|journal=Brain Connectivity|date=1 September 2019|volume=9|issue=7|pages=580–592|doi=10.1089/brain.2019.0679|pmid=31144523|pmc=6775495|ref=Hutton}}</ref> Within the dorsal attention network, the intraparietal sulcus and frontal eye fields influence the visual areas of the brain. These influencing factors allow for the orientation of attention.<ref>{{Cite journal|last1=Fox|first1=Michael D.|last2=Corbetta|first2=Maurizio|last3=Snyder|first3=Abraham Z.|last4=Vincent|first4=Justin L.|last5=Raichle|first5=Marcus E.|date=2006-06-27|title=Spontaneous neuronal activity distinguishes human dorsal and ventral attention systems|journal=Proceedings of the National Academy of Sciences|language=en|volume=103|issue=26|pages=10046–10051|doi=10.1073/pnas.0604187103|issn=0027-8424|pmid=16788060|pmc=1480402|bibcode=2006PNAS..10310046F|doi-access=free}}</ref><ref name="Vossel" /><ref name="Bailey" /><br />
<br />
<br />
*This network is involved in the voluntary, top-down deployment of attention. Within the dorsal attention network, the intraparietal sulcus and frontal eye fields influence the visual areas of the brain. These influencing factors allow for the orientation of attention.<br />
<br />
=<nowiki>= = 注意网络(背侧额顶骨) = =</nowiki>= <br />
*这个网络参与了自发的、自上而下的注意力分配。在背侧注意网络中，顶内沟和额眼影响着大脑的视觉区域。这些影响因素决定了注意力的方向。<br />
<br />
===Control (Lateral frontoparietal)===<br />
{{Main|Frontoparietal network}}<br />
* This network initiates and modulates cognitive control and comprises 18 sub-regions of the brain.<ref>{{Cite journal|last1=Scolari|first1=Miranda|last2=Seidl-Rathkopf|first2=Katharina N|last3=Kastner|first3=Sabine|date=2015-02-01|title=Functions of the human frontoparietal attention network: Evidence from neuroimaging|journal=Current Opinion in Behavioral Sciences|series=Cognitive control|volume=1|pages=32–39|doi=10.1016/j.cobeha.2014.08.003|issn=2352-1546|pmid=27398396|pmc=4936532}}</ref> There is a strong correlation between fluid intelligence and the involvement of the fronto-parietal network with other networks.<ref>{{Cite journal|last1=Marek|first1=Scott|last2=Dosenbach|first2=Nico U. F.|date=June 2018|title=The frontoparietal network: function, electrophysiology, and importance of individual precision mapping|journal=Dialogues in Clinical Neuroscience|volume=20|issue=2|pages=133–140|doi=10.31887/DCNS.2018.20.2/smarek|issn=1294-8322|pmc=6136121|pmid=30250390}}</ref><br />
<br />
<br />
*This network initiates and modulates cognitive control and comprises 18 sub-regions of the brain. There is a strong correlation between fluid intelligence and the involvement of the fronto-parietal network with other networks.<br />
<br />
=<nowiki>= = 控制网络(侧额顶骨) = =</nowiki>= <br />
*这个网络参与了认知控制的启动与调节，它包括了大脑的18个亚区。额顶网络与其他网络的参与程度和流体智力之间存在着有很强的相关性。<br />
<br />
*Versions of this network have also been called the central executive (or executive control) network and the cognitive control network.<ref name="Uddin2019" /><br />
<br />
*Versions of this network have also been called the central executive (or executive control) network and the cognitive control network.<br />
<br />
<br />
*在其它版本中，这种网络也被称为中央执行(或执行控制)网络和认知控制网络。<br />
<br />
===Sensorimotor or Somatomotor (Pericentral)===<br />
{{Main|Sensorimotor network}}<br />
*This network processes somatosensory information and coordinates motion.<ref name="Heine" /><ref name="Yeo" /><ref name="Shafiei" /><ref name="Bassett" /><ref name="Yuan" /> The [[auditory cortex]] may be included.<ref name="Uddin2019" /><ref name="Yeo" /><br />
<br />
<br />
*This network processes somatosensory information and coordinates motion. The auditory cortex may be included.<br />
<br />
===Sensorimotor or Somatomotor (Pericentral)===<br />
<br />
<br />
*This network processes somatosensory information and coordinates motion.听觉皮层可能也包括在内。<br />
<br />
===Visual (Occipital)===<br />
{{See|Visual cortex}}<br />
* This network handles visual information processing.<ref name="Yang">{{cite journal|last1=Yang|first1=Yan-li|last2=Deng|first2=Hong-xia|last3=Xing|first3=Gui-yang|last4=Xia|first4=Xiao-luan|last5=Li|first5=Hai-fang|title=Brain functional network connectivity based on a visual task: visual information processing-related brain regions are significantly activated in the task state|journal=Neural Regeneration Research|date=2015|volume=10|issue=2|pages=298–307|doi=10.4103/1673-5374.152386|pmid=25883631|pmc=4392680 }}</ref><br />
<br />
<br />
* This network handles visual information processing.<br />
<br />
= = = = = = = = = = = = = 这个网络处理视觉信息。<br />
<br />
== Other networks==<br />
Different methods and data have identified several other brain networks, many of which greatly overlap or are subsets of more well-characterized core networks.<ref name="Uddin2019" /><br />
*Limbic<ref name="Bassett" /><ref name="Yeo" /><ref name="Bailey" /><br />
*Auditory<ref name="Yuan" /><ref name="Heine" /><br />
*Right/left executive<ref name="Yuan" /><ref name="Heine" /><br />
*Cerebellar<ref name="Bell" /><ref name="Heine" /><br />
*Spatial attention<ref name="Riedl" /><ref name="Bressler" /><br />
*Language<ref name="Bressler" /><ref name="Hutton" /><br />
*Lateral visual<ref name="Yuan" /><ref name="Bell" /><ref name="Heine" /><br />
*Temporal<ref name="Yeo" /><ref name="Shafiei" /><br />
*Visual perception/imagery<ref name="Hutton" /><br />
<br />
Different methods and data have identified several other brain networks, many of which greatly overlap or are subsets of more well-characterized core networks.<br />
*Limbic<br />
*Auditory<br />
*Right/left executive<br />
*Cerebellar<br />
*Spatial attention<br />
*Language<br />
*Lateral visual<br />
*Temporal<br />
*Visual perception/imagery<br />
<br />
其他网络不同的方法和数据已经确定了其他几个大脑网络，其中许多网络极大地重叠或者是更具特色的核心网络的子集。<br />
*边缘<br />
*听觉<br />
*右/左执行<br />
*小脑<br />
*空间注意<br />
*语言<br />
*外侧视觉<br />
*颞视知觉/图像<br />
<br />
==See also==<br />
*[[Complex network]]<br />
*[[Neural network]]<br />
<br />
*Complex network<br />
*Neural network<br />
<br />
=<nowiki>= = =</nowiki>= <br />
*复杂网络<br />
*<br />
*神经网络<br />
*<br />
<br />
==References==<br />
{{reflist|30em}}<br />
<br />
{{Human connectomics}}<br />
<br />
[[Category:Neuroscience]]<br />
[[Category:Neural coding]]<br />
[[Category:Neural circuits]]<br />
[[Category:Neurophysiology]]<br />
<br />
Category:Neuroscience<br />
Category:Neural coding<br />
Category:Neural circuits<br />
Category:Neurophysiology<br />
<br />
类别: 神经科学类别: 神经编码类别: 神经回路类别: 神经生理学<br />
<br />
<noinclude><br />
<br />
<small>This page was moved from [[wikipedia:en:Large-scale brain network]]. Its edit history can be viewed at [[大规模脑网络/edithistory]]</small></noinclude><br />
<br />
[[Category:待整理页面]]</div>Shenky20https://wiki.swarma.org/index.php?title=%E5%A4%A7%E8%A7%84%E6%A8%A1%E8%84%91%E7%BD%91%E7%BB%9C&diff=29987大规模脑网络2022-04-04T08:15:18Z<p>Shenky20：</p>
<hr />
<div>此词条由神经动力学模型读书会词条梳理志愿者Shenky20翻译审校，翻译字数共800，未经专家审核，带来阅读不便，请见谅。<br />
<br />
{{Short description|Collections of brain regions}}<br />
'''Large-scale brain networks''' (also known as '''intrinsic brain networks''') are collections of widespread [[brain regions]] showing [[Resting state fMRI#Functional|functional connectivity]] by statistical analysis of the [[Functional magnetic resonance imaging|fMRI]] [[BOLD signal]]<ref name=Riedl>{{cite journal|last1=Riedl|first1=Valentin|last2=Utz|first2=Lukas|last3=Castrillón|first3=Gabriel|last4=Grimmer|first4=Timo|last5=Rauschecker|first5=Josef P.|last6=Ploner|first6=Markus|last7=Friston|first7=Karl J.|last8=Drzezga|first8=Alexander|last9=Sorg|first9=Christian|title=Metabolic connectivity mapping reveals effective connectivity in the resting human brain|journal=PNAS|date=January 12, 2016|volume=113|issue=2|pages=428–433|doi=10.1073/pnas.1513752113|pmid=26712010|pmc=4720331|bibcode=2016PNAS..113..428R|doi-access=free}}</ref> or other recording methods such as [[Electroencephalography|EEG]],<ref name=":1">{{Cite journal|last1=Foster|first1=Brett L.|last2=Parvizi|first2=Josef|date=2012-03-01|title=Resting oscillations and cross-frequency coupling in the human posteromedial cortex|journal=NeuroImage|volume=60|issue=1|pages=384–391|doi=10.1016/j.neuroimage.2011.12.019|pmid=22227048|issn=1053-8119|pmc=3596417}}</ref> [[Positron emission tomography|PET]]<ref name=":2">{{Cite journal|last1=Buckner|first1=Randy L.|last2=Andrews‐Hanna|first2=Jessica R.|last3=Schacter|first3=Daniel L.|date=2008|title=The Brain's Default Network|journal=Annals of the New York Academy of Sciences|language=en|volume=1124|issue=1|pages=1–38|doi=10.1196/annals.1440.011|pmid=18400922|issn=1749-6632|bibcode=2008NYASA1124....1B|s2cid=3167595}}</ref> and [[Magnetoencephalography|MEG]].<ref name=":3">{{Cite journal|last1=Morris|first1=Peter G.|last2=Smith|first2=Stephen M.|last3=Barnes|first3=Gareth R.|last4=Stephenson|first4=Mary C.|last5=Hale|first5=Joanne R.|last6=Price|first6=Darren|last7=Luckhoo|first7=Henry|last8=Woolrich|first8=Mark|last9=Brookes|first9=Matthew J.|date=2011-10-04|title=Investigating the electrophysiological basis of resting state networks using magnetoencephalography|journal=Proceedings of the National Academy of Sciences|language=en|volume=108|issue=40|pages=16783–16788|doi=10.1073/pnas.1112685108|issn=0027-8424|pmid=21930901|pmc=3189080|bibcode=2011PNAS..10816783B|doi-access=free}}</ref> An emerging paradigm in neuroscience is that cognitive tasks are performed not by individual brain regions working in isolation but by networks consisting of several discrete brain regions that are said to be "functionally connected". Functional connectivity networks may be found using algorithms such as [[cluster analysis]], spatial [[independent component analysis]] (ICA), seed based, and others.<ref name="Petersen">{{cite journal|last1=Petersen|first1=Steven|last2=Sporns|first2=Olaf|title=Brain Networks and Cognitive Architectures|journal=Neuron|date=October 2015|volume=88|issue=1|pages=207–219|doi=10.1016/j.neuron.2015.09.027|pmid=26447582|pmc=4598639 }}</ref> Synchronized brain regions may also be identified using long-range synchronization of the EEG, MEG, or other dynamic brain signals.<ref name=Bressler>{{cite journal|last1=Bressler|first1=Steven L.|last2=Menon|first2=Vinod|s2cid=5967761|title=Large scale brain networks in cognition: emerging methods and principles|journal=Trends in Cognitive Sciences|date=June 2010|volume=14|issue=6|pages=233–290|doi=10.1016/j.tics.2010.04.004|url=http://www.cell.com/trends/cognitive-sciences/issue?pii=S1364-6613(10)X0005-5|accessdate=24 January 2016|pmid=20493761}}</ref><br />
<br />
<br />
Large-scale brain networks (also known as intrinsic brain networks) are collections of widespread brain regions showing functional connectivity by statistical analysis of the fMRI BOLD signal or other recording methods such as EEG, PET and MEG. An emerging paradigm in neuroscience is that cognitive tasks are performed not by individual brain regions working in isolation but by networks consisting of several discrete brain regions that are said to be "functionally connected". Functional connectivity networks may be found using algorithms such as cluster analysis, spatial independent component analysis (ICA), seed based, and others. Synchronized brain regions may also be identified using long-range synchronization of the EEG, MEG, or other dynamic brain signals.<br />
<br />
'''<font color="#ff8000">大规模脑网络Large-scale brain networks</font>'''(也称为'''<font color="#ff8000">内在大脑网络Intrinsic brain networks</font>''')是在对基于'''<font color="#ff8000">血氧水平依赖效应BOLD</font>'''的'''<font color="#ff8000">功能性磁共振成像fMRI</font>'''信号<ref name="Riedl" />的统计分析或其他记录方法(如'''<font color="#ff8000">脑电图EEG<ref name=":1" /></font>'''、'''<font color="#ff8000">正电子发射断层扫描技术PET<ref name=":2" /></font>'''和'''<font color="#ff8000">脑磁图MEG<ref name=":3" /></font>''')中，表现出'''<font color="#ff8000">功能连接Functional connectivity</font>'''的'''<font color="#ff8000">脑区Brain regions</font>'''的集合。根据神经科学中一个新出现的范式，认知任务不是由单个脑区独立执行的，而是由几个互不相连的脑区“功能连接”组成的网络执行的。功能连接网络可以通过'''<font color="#ff8000">数据聚类Cluster analysis</font>'''、空间'''<font color="#ff8000">独立元素分析ICA</font>'''、种子点方法等算法来发现。<ref name="Petersen" />同步的脑区也可以用脑电图、脑磁图或其他动态脑信号的远程同步来识别。<ref name="Bressler" /><br />
<br />
The set of identified brain areas that are linked together in a large-scale network varies with cognitive function.<ref name="Bressler2">{{cite journal|last1=Bressler|first1=Steven L.|title=Neurocognitive networks|journal=Scholarpedia|volume=3|issue=2|pages=1567|doi=10.4249/scholarpedia.1567|year=2008|bibcode=2008SchpJ...3.1567B|doi-access=free}}</ref> When the cognitive state is not explicit (i.e., the subject is at "rest"), the large-scale brain network is a [[Resting state fMRI|resting state]] network (RSN). As a physical system with graph-like properties,<ref name="Bressler" /> a large-scale brain network has both nodes and edges and cannot be identified simply by the co-activation of brain areas. In recent decades, the analysis of brain networks was made feasible by advances in imaging techniques as well as new tools from [[graph theory]] and [[Dynamical systems theory|dynamical systems]].<br />
<br />
The set of identified brain areas that are linked together in a large-scale network varies with cognitive function. When the cognitive state is not explicit (i.e., the subject is at "rest"), the large-scale brain network is a resting state network (RSN). As a physical system with graph-like properties, a large-scale brain network has both nodes and edges and cannot be identified simply by the co-activation of brain areas. In recent decades, the analysis of brain networks was made feasible by advances in imaging techniques as well as new tools from graph theory and dynamical systems.<br />
<br />
大规模脑网络中,连接在一起的脑区集合因认知功能的不同而不同。<ref name="Bressler2" />当认知状态不明确(即主体处于“静止”状态)时，大规模脑网络是一个'''<font color="#ff8000">静息状态Resting state</font>'''网络(RSN)。作为一个具有图形特征的物理系统，<ref name="Bressler" />大规模脑网络既有节点又有边，不能简单地通过脑区的共同激活来识别。近几十年来，成像技术不断进步，同时'''<font color="#ff8000">图论Graph theory</font>'''、'''<font color="#ff8000">动力学系统Dynamical systems</font>'''领域出现了新的技术手段，这使得脑网络分析变得可行。<br />
<br />
Large-scale brain networks are identified by their function and provide a coherent framework for understanding [[cognition]] by offering a neural model of how different cognitive functions emerge when different sets of brain regions join together as self-organized coalitions. The number and composition of the coalitions will vary with the algorithm and parameters used to identify them.<ref name="Yeo" /><ref name=":4">{{cite journal|last1=Abou Elseoud|first1=Ahmed|last2=Littow|first2=Harri|last3=Remes|first3=Jukka|last4=Starck|first4=Tuomo|last5=Nikkinen|first5=Juha|last6=Nissilä|first6=Juuso|last7=Timonen|first7=Markku|last8=Tervonen|first8=Osmo|last9=Kiviniemi1|first9=Vesa|title=Group-ICA Model Order Highlights Patterns of Functional Brain Connectivity|journal= Frontiers in Systems Neuroscience|date=2011-06-03|volume=5|pages=37|doi=10.3389/fnsys.2011.00037|pmid=21687724|pmc=3109774 |doi-access=free}}</ref> In one model, there is only the [[default mode network]] and the [[task-positive network]], but most current analyses show several networks, from a small handful to 17.<ref name="Yeo" /> The most common and stable networks are enumerated below. The regions participating in a functional network may be dynamically reconfigured.<ref name="Petersen" /><ref name="Bassett">{{cite journal|last1=Bassett|first1=Daniella|last2=Bertolero|first2=Max|title=How Matter Becomes Mind|journal=Scientific American|date=July 2019|volume=321|issue=1|page=32|url=https://www.scientificamerican.com/|accessdate=23 June 2019 }}</ref><br />
<br />
Large-scale brain networks are identified by their function and provide a coherent framework for understanding cognition by offering a neural model of how different cognitive functions emerge when different sets of brain regions join together as self-organized coalitions. The number and composition of the coalitions will vary with the algorithm and parameters used to identify them. In one model, there is only the default mode network and the task-positive network, but most current analyses show several networks, from a small handful to 17. The most common and stable networks are enumerated below. The regions participating in a functional network may be dynamically reconfigured.<br />
<br />
大规模脑网络是通过其功能来进行识别的。通过研究大规模脑网络建立神经模型，对不同脑区组合所形成的自组织联合体如何实现不同的'''<font color="#ff8000">认知Cognition</font>'''功能进行解释，就能够为认知理解提供一个连贯的框架。识别算法和参数的不同会导致所识别出的上述联合体的数量和组成有所不同。<ref name="Yeo" /><ref name=":4" />一个模型理论认为，符合上述条件的神经模型只包含'''<font color="#ff8000">默认模式网络Default mode network</font>'''和'''<font color="#ff8000">任务激活网络Task-positive network</font>'''，但目前'''<font color="#32CD32">大多数分析理论都包括从几个到17个不等的网络。<ref name="Yeo" /></font>'''下面列举了最常见且稳定的网络。'''<font color="#32CD32"> 人脑</font>'''可以动态地重新配置参与功能网络的脑区。<ref name="Petersen" /><ref name="Bassett" /> <br />
<br />
Disruptions in activity in various networks have been implicated in neuropsychiatric disorders such as [[Depression (mood)|depression]], [[Alzheimer's disease|Alzheimer's]], [[Autism-spectrum disorder|autism spectrum disorder]], [[schizophrenia]], [[ADHD]]<ref name=":5">{{cite journal |last1=Griffiths |first1=Kristi R. |last2=Braund |first2=Taylor A. |last3=Kohn |first3=Michael R. |last4=Clarke |first4=Simon |last5=Williams |first5=Leanne M. |last6=Korgaonkar |first6=Mayuresh S. |title=Structural brain network topology underpinning ADHD and response to methylphenidate treatment |journal=Translational Psychiatry |date=2 March 2021 |volume=11 |issue=1 |pages=1–9 |doi=10.1038/s41398-021-01278-x | pmc=7925571 |pmid=33654073 |url=https://www.nature.com/articles/s41398-021-01278-x#citeas |access-date=16 November 2021}}</ref> and [[bipolar disorder]].<ref name=":6">{{Cite journal|url=https://www.researchgate.net/publication/51639686|title=Large-scale brain networks and psychopathology: A unifying triple network model|last=Menon|first=Vinod|s2cid=26653572|journal=Trends in Cognitive Sciences|date=2011-09-09|volume=15|issue=10|pages=483–506|doi=10.1016/j.tics.2011.08.003|pmid=21908230}}</ref><br />
<br />
Disruptions in activity in various networks have been implicated in neuropsychiatric disorders such as depression, Alzheimer's, autism spectrum disorder, schizophrenia, ADHD and bipolar disorder.<br />
<br />
脑网络活动的中断与诸多神经精神疾病密切相关，如'''<font color="#ff8000">抑郁症Depression</font>'''、'''<font color="#ff8000">老年痴呆症Alzheimer's</font>'''、'''<font color="#ff8000">自闭症谱系障碍Autism spectrum disorder</font>'''、'''<font color="#ff8000">精神分裂症Schizophrenia</font>'''、'''<font color="#ff8000">多动症ADHD<ref name=":5" /></font>'''和'''<font color="#ff8000">躁郁症Bipolar disorder<ref name=":6" /></font>'''。<br />
<br />
== Core networks==<br />
[[File:Heine2012x3010.png|thumb|An example that identified 10 large-scale brain networks from [[resting state fMRI]] activity through [[independent component analysis]].<ref name="Heine" />|链接=Special:FilePath/Heine2012x3010.png]]<br />
Because brain networks can be identified at various different resolutions and with various different neurobiological properties, there is no such thing as a universal atlas of brain networks that fits all circumstances.<ref name=":7">{{cite journal|last1=Eickhoff|first1=SB|last2=Yeo|first2=BTT|last3=Genon|first3=S|title=Imaging-based parcellations of the human brain.|journal=Nature Reviews. Neuroscience|date=November 2018|volume=19|issue=11|pages=672–686|doi=10.1038/s41583-018-0071-7|pmid=30305712|s2cid=52954265|url=http://juser.fz-juelich.de/record/856633/files/Eickhoff_Yeo_Genon_NRN_MainManuscriptInclFigures.pdf}}</ref> While acknowledging this problem, Uddin, Yeo, and Spreng proposed in 2019<ref name="Uddin2019">{{cite journal|last1=Uddin|first1=LQ|last2=Yeo|first2=BTT|last3=Spreng|first3=RN|title=Towards a Universal Taxonomy of Macro-scale Functional Human Brain Networks.|journal=Brain Topography|date=November 2019|volume=32|issue=6|pages=926–942|doi=10.1007/s10548-019-00744-6|pmid=31707621|pmc=7325607}}</ref> that the following six networks should be defined as core networks based on converging evidences from multiple studies<ref name=":8">{{cite journal|last1=Doucet|first1=GE|last2=Lee|first2=WH|last3=Frangou|first3=S|title=Evaluation of the spatial variability in the major resting-state networks across human brain functional atlases.|journal=Human Brain Mapping|date=2019-10-15|volume=40|issue=15|pages=4577–4587|doi=10.1002/hbm.24722|pmid=31322303|pmc=6771873}}</ref><ref name="Yeo" /><ref name=":9">{{cite journal|last1=Smith|first1=SM|last2=Fox|first2=PT|last3=Miller|first3=KL|last4=Glahn|first4=DC|last5=Fox|first5=PM|last6=Mackay|first6=CE|last7=Filippini|first7=N|last8=Watkins|first8=KE|last9=Toro|first9=R|last10=Laird|first10=AR|last11=Beckmann|first11=CF|title=Correspondence of the brain's functional architecture during activation and rest.|journal=Proceedings of the National Academy of Sciences of the United States of America|date=2009-08-04|volume=106|issue=31|pages=13040–5|doi=10.1073/pnas.0905267106|pmid=19620724|pmc=2722273|bibcode=2009PNAS..10613040S|doi-access=free}}</ref> to facilitate communication between researchers.<br />
<br />
【图1：An example that identified 10 large-scale brain networks from resting state fMRI activity through independent component analysis.<ref name="Heine" />这是一个通过独立元素分析方法从静息状态脑功能性磁共振成像信号中分辨出10个大规模脑网络的例子。<ref name="Heine" />】<br />
<br />
Because brain networks can be identified at various different resolutions and with various different neurobiological properties, there is no such thing as a universal atlas of brain networks that fits all circumstances. While acknowledging this problem, Uddin, Yeo, and Spreng proposed in 2019 that the following six networks should be defined as core networks based on converging evidences from multiple studies to facilitate communication between researchers.<br />
<br />
因为脑网络可以在不同分辨率条件下、使用不同的神经生物学特性来识别，所以没有能够适用于所有情况的通用脑网络图谱。<ref name=":7" />在承认这个问题的同时，Uddin，Yeo，和Spreng在2019年提出<ref name="Uddin2019" />，综合考虑来自多个研究的证据<ref name=":8" /><ref name="Yeo" /><ref name=":9" />，以下六个网络应该被定义为核心网络，以促进研究人员之间的交流。<br />
<br />
===Default Mode (Medial frontoparietal)===<br />
{{Main|Default mode network}}<br />
*The default mode network is active when an individual is awake and at rest. It preferentially activates when individuals focus on internally-oriented tasks such as daydreaming, envisioning the future, retrieving memories, and [[theory of mind]]. It is negatively correlated with brain systems that focus on external visual signals. It is the most widely researched network.<ref name="Bressler" /><ref name="Bassett" /><ref>{{Cite journal|date=2012-08-15|title=The serendipitous discovery of the brain's default network|journal=NeuroImage|language=en|volume=62|issue=2|pages=1137–1145|doi=10.1016/j.neuroimage.2011.10.035|pmid=22037421|issn=1053-8119|last1=Buckner|first1=Randy L.|s2cid=9880586}}</ref><ref name="Riedl" /><ref name="Yuan"><br />
{{cite journal|last1=Yuan|first1=Rui|last2=Di|first2=Xin|last3=Taylor|first3=Paul A.|last4=Gohel|first4=Suril|last5=Tsai|first5=Yuan-Hsiung|last6=Biswal|first6=Bharat B.|title=Functional topography of the thalamocortical system in human|journal=Brain Structure and Function|date=30 April 2015|doi=10.1007/s00429-015-1018-7|pmid=25924563|pmc=6363530|volume=221|issue=4|pages=1971–1984}}</ref><ref name="Bell">{{cite journal|last1=Bell|first1=Peter T.|last2=Shine|first2=James M.|title=Estimating Large-Scale Network Convergence in the Human Functional Connectome|journal=Brain Connectivity|date=2015-11-09|volume=5|issue=9|doi=10.1089/brain.2015.0348|pmid=26005099|pages=565–74}}</ref><ref name="Heine">{{cite journal|last1=Heine|first1=Lizette|last2=Soddu|first2=Andrea|last3=Gomez|first3=Francisco|last4=Vanhaudenhuyse|first4=Audrey|last5=Tshibanda|first5=Luaba|last6=Thonnard|first6=Marie|last7=Charland-Verville|first7=Vanessa|last8=Kirsch|first8=Murielle|last9=Laureys|first9=Steven|last10=Demertzi|first10=Athena|title=Resting state networks and consciousness. Alterations of multiple resting state network connectivity in physiological, pharmacological and pathological consciousness states.|journal=Frontiers in Psychology|date=2012|volume=3|pages=295|doi=10.3389/fpsyg.2012.00295|pmid=22969735|pmc=3427917|doi-access=free}}</ref><ref name="Yeo">{{cite journal|last1=Yeo|first1=B. T. Thomas|last2=Krienen|first2=Fenna M.|last3=Sepulcre|first3=Jorge|last4=Sabuncu|first4=Mert R.|last5=Lashkari|first5=Danial|last6=Hollinshead|first6=Marisa|last7=Roffman|first7=Joshua L.|last8=Smoller|first8=Jordan W.|last9=Zöllei|first9=Lilla|last10=Polimeni|first10=Jonathan R.|last11=Fischl|first11=Bruce|last12=Liu|first12=Hesheng|last13=Buckner|first13=Randy L.|title=The organization of the human cerebral cortex estimated by intrinsic functional connectivity|journal=Journal of Neurophysiology|date=2011-09-01|volume=106|issue=3|pages=1125–1165|doi=10.1152/jn.00338.2011|pmid=21653723|pmc=3174820|bibcode=2011NatSD...2E0031H }}</ref><ref name="Shafiei">{{cite journal|last1=Shafiei|first1=Golia|last2=Zeighami|first2=Yashar|last3=Clark|first3=Crystal A.|last4=Coull|first4=Jennifer T.|last5=Nagano-Saito|first5=Atsuko|last6=Leyton|first6=Marco|last7=Dagher|first7=Alain|last8=Mišić|first8=Bratislav|title=Dopamine Signaling Modulates the Stability and Integration of Intrinsic Brain Networks|journal=Cerebral Cortex|date=2018-10-01|volume=29|issue=1|pages=397–409|doi=10.1093/cercor/bhy264|pmid=30357316|pmc=6294404 }}</ref><ref name="Bailey">{{cite journal|last1=Bailey|first1=Stephen K.|last2=Aboud|first2=Katherine S.|last3=Nguyen|first3=Tin Q.|last4=Cutting|first4=Laurie E.|title=Applying a network framework to the neurobiology of reading and dyslexia|journal=Journal of Neurodevelopmental Disorders|date=13 December 2018|volume=10|issue=1|page=37|doi=10.1186/s11689-018-9251-z|pmid=30541433|pmc=6291929 }}</ref><br />
<br />
<br />
* The default mode network is active when an individual is awake and at rest. It preferentially activates when individuals focus on internally-oriented tasks such as daydreaming, envisioning the future, retrieving memories, and theory of mind. It is negatively correlated with brain systems that focus on external visual signals. It is the most widely researched network.<br />
<br />
<br />
=<nowiki>= = 默认模式(Medial frontoparietal) = =</nowiki>= <br />
*默认模式网络在个人清醒和休息时处于活动状态。当个体专注于面向内部的任务时，比如做白日梦、展望未来、回忆和心理理论，它就会优先激活。它与专注于外部视觉信号的大脑系统负相关。它是研究最广泛的网络。<br />
<br />
===Salience (Midcingulo-Insular)===<br />
{{Main|Salience network}}<br />
*The salience network consists of several structures, including the anterior (bilateral) insula, dorsal anterior cingulate cortex, and three subcortical structures which are the ventral striatum, substantia nigra/ventral tegmental region.<ref>{{Cite journal|last1=Steimke|first1=Rosa|last2=Nomi|first2=Jason S.|last3=Calhoun|first3=Vince D.|last4=Stelzel|first4=Christine|last5=Paschke|first5=Lena M.|last6=Gaschler|first6=Robert|last7=Goschke|first7=Thomas|last8=Walter|first8=Henrik|last9=Uddin|first9=Lucina Q.|date=2017-12-01|title=Salience network dynamics underlying successful resistance of temptation|journal=Social Cognitive and Affective Neuroscience|language=en|volume=12|issue=12|pages=1928–1939|doi=10.1093/scan/nsx123|pmid=29048582|pmc=5716209|issn=1749-5016|doi-access=free}}</ref><ref name=":0">{{Citation|last=Menon|first=V.|title=Brain Mapping|chapter=Salience Network|date=2015-01-01|chapter-url=http://www.sciencedirect.com/science/article/pii/B978012397025100052X|pages=597–611|editor-last=Toga|editor-first=Arthur W.|publisher=Academic Press|isbn=978-0-12-397316-0|access-date=2019-12-08|doi=10.1016/B978-0-12-397025-1.00052-X}}</ref> It plays the key role of monitoring the [[Salience (neuroscience)|salience]] of external inputs and internal brain events.<ref name="Riedl" /><ref name="Bressler" /><ref name="Bassett" /><ref name="Yuan" /><ref name="Heine" /><ref name="Yeo" /><ref name="Shafiei" /> Specifically, it aids in directing attention by identifying important biological and cognitive events.<ref name=":0" /><ref name="Bailey" /><br />
*This network includes the ventral attention network, which primarily includes the [[temporoparietal junction]] and the ventral [[frontal cortex]] of the right hemisphere.<ref name="Uddin2019" /><ref name="Vossel" /> These areas respond when behaviorally relevant stimuli occur unexpectedly.<ref name="Vossel" /> The ventral attention network is inhibited during focused attention in which top-down processing is being used, such as when visually searching for something. This response may prevent goal-driven attention from being distracted by non-relevant stimuli. It becomes active again when the target or relevant information about the target is found.<ref name="Vossel" /><ref>{{Cite journal|last1=Shulman|first1=Gordon L.|last2=McAvoy|first2=Mark P.|last3=Cowan|first3=Melanie C.|last4=Astafiev|first4=Serguei V.|last5=Tansy|first5=Aaron P.|last6=d'Avossa|first6=Giovanni|last7=Corbetta|first7=Maurizio|date=2003-11-01|title=Quantitative Analysis of Attention and Detection Signals During Visual Search|journal=Journal of Neurophysiology|volume=90|issue=5|pages=3384–3397|doi=10.1152/jn.00343.2003|pmid=12917383|issn=0022-3077}}</ref><br />
<br />
<br />
*The salience network consists of several structures, including the anterior (bilateral) insula, dorsal anterior cingulate cortex, and three subcortical structures which are the ventral striatum, substantia nigra/ventral tegmental region. It plays the key role of monitoring the salience of external inputs and internal brain events. Specifically, it aids in directing attention by identifying important biological and cognitive events.<br />
*This network includes the ventral attention network, which primarily includes the temporoparietal junction and the ventral frontal cortex of the right hemisphere. These areas respond when behaviorally relevant stimuli occur unexpectedly. The ventral attention network is inhibited during focused attention in which top-down processing is being used, such as when visually searching for something. This response may prevent goal-driven attention from being distracted by non-relevant stimuli. It becomes active again when the target or relevant information about the target is found.<br />
<br />
突显网络由前(双)岛、前扣带回皮层和腹侧纹状体、黑质/腹侧被盖区3个皮层下结构组成。它起着监测外部输入和内部脑事件的显著性的关键作用。具体来说，它通过识别重要的生物学和认知事件来帮助引导注意力。<br />
*这个网络包括腹侧注意网络，主要包括右半球的颞顶联合区和腹侧额叶皮层。当行为相关的刺激意外发生时，这些区域会做出反应。腹侧注意网络在使用自上而下加工的集中注意过程中被抑制，例如在视觉搜索某物时。这种反应可以防止目标驱动的注意力被非相关的刺激分散。当找到目标或关于目标的相关信息时，它再次激活。<br />
<br />
===Attention (Dorsal frontoparietal)===<br />
{{Main|Dorsal attention network}}<br />
*This network is involved in the voluntary, top-down deployment of attention.<ref name="Riedl" /><ref name="Yuan" /><ref name="Bell" /><ref name="Yeo" /><ref name="Shafiei" /><ref name="Vossel">{{cite journal|last1=Vossel|first1=Simone|last2=Geng|first2=Joy J.|last3=Fink|first3=Gereon R.|title=Dorsal and Ventral Attention Systems: Distinct Neural Circuits but Collaborative Roles|journal=The Neuroscientist|date=2014|volume=20|issue=2|pages=150–159|doi=10.1177/1073858413494269|pmid=23835449|pmc=4107817}}</ref><ref name="Hutton">{{cite journal|last1=Hutton|first1=John S.|last2=Dudley|first2=Jonathan|last3=Horowitz-Kraus|first3=Tzipi|last4=DeWitt|first4=Tom|last5=Holland|first5=Scott K.|title=Functional Connectivity of Attention, Visual, and Language Networks During Audio, Illustrated, and Animated Stories in Preschool-Age Children|journal=Brain Connectivity|date=1 September 2019|volume=9|issue=7|pages=580–592|doi=10.1089/brain.2019.0679|pmid=31144523|pmc=6775495|ref=Hutton}}</ref> Within the dorsal attention network, the intraparietal sulcus and frontal eye fields influence the visual areas of the brain. These influencing factors allow for the orientation of attention.<ref>{{Cite journal|last1=Fox|first1=Michael D.|last2=Corbetta|first2=Maurizio|last3=Snyder|first3=Abraham Z.|last4=Vincent|first4=Justin L.|last5=Raichle|first5=Marcus E.|date=2006-06-27|title=Spontaneous neuronal activity distinguishes human dorsal and ventral attention systems|journal=Proceedings of the National Academy of Sciences|language=en|volume=103|issue=26|pages=10046–10051|doi=10.1073/pnas.0604187103|issn=0027-8424|pmid=16788060|pmc=1480402|bibcode=2006PNAS..10310046F|doi-access=free}}</ref><ref name="Vossel" /><ref name="Bailey" /><br />
<br />
<br />
*This network is involved in the voluntary, top-down deployment of attention. Within the dorsal attention network, the intraparietal sulcus and frontal eye fields influence the visual areas of the brain. These influencing factors allow for the orientation of attention.<br />
<br />
=<nowiki>= = 注意力(背侧额顶骨) = =</nowiki>= <br />
*这个网络参与了自发的、自上而下的注意力分配。在背侧注意网络中，顶内沟和额眼影响大脑的视觉区域。这些影响因素决定了注意力的方向。<br />
<br />
===Control (Lateral frontoparietal)===<br />
{{Main|Frontoparietal network}}<br />
*This network initiates and modulates cognitive control and comprises 18 sub-regions of the brain.<ref>{{Cite journal|last1=Scolari|first1=Miranda|last2=Seidl-Rathkopf|first2=Katharina N|last3=Kastner|first3=Sabine|date=2015-02-01|title=Functions of the human frontoparietal attention network: Evidence from neuroimaging|journal=Current Opinion in Behavioral Sciences|series=Cognitive control|volume=1|pages=32–39|doi=10.1016/j.cobeha.2014.08.003|issn=2352-1546|pmid=27398396|pmc=4936532}}</ref> There is a strong correlation between fluid intelligence and the involvement of the fronto-parietal network with other networks.<ref>{{Cite journal|last1=Marek|first1=Scott|last2=Dosenbach|first2=Nico U. F.|date=June 2018|title=The frontoparietal network: function, electrophysiology, and importance of individual precision mapping|journal=Dialogues in Clinical Neuroscience|volume=20|issue=2|pages=133–140|doi=10.31887/DCNS.2018.20.2/smarek|issn=1294-8322|pmc=6136121|pmid=30250390}}</ref><br />
<br />
<br />
*This network initiates and modulates cognitive control and comprises 18 sub-regions of the brain. There is a strong correlation between fluid intelligence and the involvement of the fronto-parietal network with other networks.<br />
<br />
=<nowiki>= = 控制(侧额顶骨) = =</nowiki>= <br />
*这个网络启动和调节认知控制，包括大脑的18个亚区。在流体智力和额顶网络与其他网络的参与之间有很强的相关性。<br />
<br />
*Versions of this network have also been called the central executive (or executive control) network and the cognitive control network.<ref name="Uddin2019" /><br />
<br />
*Versions of this network have also been called the central executive (or executive control) network and the cognitive control network.<br />
<br />
<br />
*这种网络的版本也被称为中央执行(或执行控制)网络和认知控制网络。<br />
<br />
===Sensorimotor or Somatomotor (Pericentral)===<br />
{{Main|Sensorimotor network}}<br />
*This network processes somatosensory information and coordinates motion.<ref name="Heine" /><ref name="Yeo" /><ref name="Shafiei" /><ref name="Bassett" /><ref name="Yuan" /> The [[auditory cortex]] may be included.<ref name="Uddin2019" /><ref name="Yeo" /><br />
<br />
<br />
*This network processes somatosensory information and coordinates motion. The auditory cortex may be included.<br />
<br />
===Sensorimotor or Somatomotor (Pericentral)===<br />
<br />
<br />
*This network processes somatosensory information and coordinates motion.听觉皮层可能也包括在内。<br />
<br />
===Visual (Occipital)===<br />
{{See|Visual cortex}}<br />
*This network handles visual information processing.<ref name="Yang">{{cite journal|last1=Yang|first1=Yan-li|last2=Deng|first2=Hong-xia|last3=Xing|first3=Gui-yang|last4=Xia|first4=Xiao-luan|last5=Li|first5=Hai-fang|title=Brain functional network connectivity based on a visual task: visual information processing-related brain regions are significantly activated in the task state|journal=Neural Regeneration Research|date=2015|volume=10|issue=2|pages=298–307|doi=10.4103/1673-5374.152386|pmid=25883631|pmc=4392680 }}</ref><br />
<br />
<br />
*This network handles visual information processing.<br />
<br />
= = = = = = = = = = = = = 这个网络处理视觉信息。<br />
<br />
==Other networks==<br />
Different methods and data have identified several other brain networks, many of which greatly overlap or are subsets of more well-characterized core networks.<ref name="Uddin2019" /><br />
*Limbic<ref name="Bassett" /><ref name="Yeo" /><ref name="Bailey" /><br />
*Auditory<ref name="Yuan" /><ref name="Heine" /><br />
*Right/left executive<ref name="Yuan" /><ref name="Heine" /><br />
*Cerebellar<ref name="Bell" /><ref name="Heine" /><br />
*Spatial attention<ref name="Riedl" /><ref name="Bressler" /><br />
*Language<ref name="Bressler" /><ref name="Hutton" /><br />
*Lateral visual<ref name="Yuan" /><ref name="Bell" /><ref name="Heine" /><br />
*Temporal<ref name="Yeo" /><ref name="Shafiei" /><br />
*Visual perception/imagery<ref name="Hutton" /><br />
<br />
Different methods and data have identified several other brain networks, many of which greatly overlap or are subsets of more well-characterized core networks.<br />
*Limbic<br />
*Auditory<br />
*Right/left executive<br />
*Cerebellar<br />
*Spatial attention<br />
*Language<br />
*Lateral visual<br />
*Temporal<br />
*Visual perception/imagery<br />
<br />
其他网络不同的方法和数据已经确定了其他几个大脑网络，其中许多网络极大地重叠或者是更具特色的核心网络的子集。<br />
*边缘<br />
*听觉<br />
*右/左执行<br />
*小脑<br />
*空间注意<br />
*语言<br />
*外侧视觉<br />
*颞视知觉/图像<br />
<br />
==See also==<br />
*[[Complex network]]<br />
*[[Neural network]]<br />
<br />
*Complex network<br />
*Neural network<br />
<br />
=<nowiki>= = =</nowiki>= <br />
*复杂网络<br />
*<br />
*神经网络<br />
*<br />
<br />
==References==<br />
{{reflist|30em}}<br />
<br />
{{Human connectomics}}<br />
<br />
[[Category:Neuroscience]]<br />
[[Category:Neural coding]]<br />
[[Category:Neural circuits]]<br />
[[Category:Neurophysiology]]<br />
<br />
Category:Neuroscience<br />
Category:Neural coding<br />
Category:Neural circuits<br />
Category:Neurophysiology<br />
<br />
类别: 神经科学类别: 神经编码类别: 神经回路类别: 神经生理学<br />
<br />
<noinclude><br />
<br />
<small>This page was moved from [[wikipedia:en:Large-scale brain network]]. Its edit history can be viewed at [[大规模脑网络/edithistory]]</small></noinclude><br />
<br />
[[Category:待整理页面]]</div>Shenky20https://wiki.swarma.org/index.php?title=%E5%A4%A7%E8%A7%84%E6%A8%A1%E8%84%91%E7%BD%91%E7%BB%9C&diff=29986大规模脑网络2022-04-04T08:03:03Z<p>Shenky20：/* Core networks */</p>
<hr />
<div>此词条由神经动力学模型读书会词条梳理志愿者Shenky20翻译审校，翻译字数共800，未经专家审核，带来阅读不便，请见谅。<br />
<br />
{{Short description|Collections of brain regions}}<br />
'''Large-scale brain networks''' (also known as '''intrinsic brain networks''') are collections of widespread [[brain regions]] showing [[Resting state fMRI#Functional|functional connectivity]] by statistical analysis of the [[Functional magnetic resonance imaging|fMRI]] [[BOLD signal]]<ref name=Riedl>{{cite journal|last1=Riedl|first1=Valentin|last2=Utz|first2=Lukas|last3=Castrillón|first3=Gabriel|last4=Grimmer|first4=Timo|last5=Rauschecker|first5=Josef P.|last6=Ploner|first6=Markus|last7=Friston|first7=Karl J.|last8=Drzezga|first8=Alexander|last9=Sorg|first9=Christian|title=Metabolic connectivity mapping reveals effective connectivity in the resting human brain|journal=PNAS|date=January 12, 2016|volume=113|issue=2|pages=428–433|doi=10.1073/pnas.1513752113|pmid=26712010|pmc=4720331|bibcode=2016PNAS..113..428R|doi-access=free}}</ref> or other recording methods such as [[Electroencephalography|EEG]],<ref name=":1">{{Cite journal|last1=Foster|first1=Brett L.|last2=Parvizi|first2=Josef|date=2012-03-01|title=Resting oscillations and cross-frequency coupling in the human posteromedial cortex|journal=NeuroImage|volume=60|issue=1|pages=384–391|doi=10.1016/j.neuroimage.2011.12.019|pmid=22227048|issn=1053-8119|pmc=3596417}}</ref> [[Positron emission tomography|PET]]<ref name=":2">{{Cite journal|last1=Buckner|first1=Randy L.|last2=Andrews‐Hanna|first2=Jessica R.|last3=Schacter|first3=Daniel L.|date=2008|title=The Brain's Default Network|journal=Annals of the New York Academy of Sciences|language=en|volume=1124|issue=1|pages=1–38|doi=10.1196/annals.1440.011|pmid=18400922|issn=1749-6632|bibcode=2008NYASA1124....1B|s2cid=3167595}}</ref> and [[Magnetoencephalography|MEG]].<ref name=":3">{{Cite journal|last1=Morris|first1=Peter G.|last2=Smith|first2=Stephen M.|last3=Barnes|first3=Gareth R.|last4=Stephenson|first4=Mary C.|last5=Hale|first5=Joanne R.|last6=Price|first6=Darren|last7=Luckhoo|first7=Henry|last8=Woolrich|first8=Mark|last9=Brookes|first9=Matthew J.|date=2011-10-04|title=Investigating the electrophysiological basis of resting state networks using magnetoencephalography|journal=Proceedings of the National Academy of Sciences|language=en|volume=108|issue=40|pages=16783–16788|doi=10.1073/pnas.1112685108|issn=0027-8424|pmid=21930901|pmc=3189080|bibcode=2011PNAS..10816783B|doi-access=free}}</ref> An emerging paradigm in neuroscience is that cognitive tasks are performed not by individual brain regions working in isolation but by networks consisting of several discrete brain regions that are said to be "functionally connected". Functional connectivity networks may be found using algorithms such as [[cluster analysis]], spatial [[independent component analysis]] (ICA), seed based, and others.<ref name="Petersen">{{cite journal|last1=Petersen|first1=Steven|last2=Sporns|first2=Olaf|title=Brain Networks and Cognitive Architectures|journal=Neuron|date=October 2015|volume=88|issue=1|pages=207–219|doi=10.1016/j.neuron.2015.09.027|pmid=26447582|pmc=4598639 }}</ref> Synchronized brain regions may also be identified using long-range synchronization of the EEG, MEG, or other dynamic brain signals.<ref name=Bressler>{{cite journal|last1=Bressler|first1=Steven L.|last2=Menon|first2=Vinod|s2cid=5967761|title=Large scale brain networks in cognition: emerging methods and principles|journal=Trends in Cognitive Sciences|date=June 2010|volume=14|issue=6|pages=233–290|doi=10.1016/j.tics.2010.04.004|url=http://www.cell.com/trends/cognitive-sciences/issue?pii=S1364-6613(10)X0005-5|accessdate=24 January 2016|pmid=20493761}}</ref><br />
<br />
<br />
Large-scale brain networks (also known as intrinsic brain networks) are collections of widespread brain regions showing functional connectivity by statistical analysis of the fMRI BOLD signal or other recording methods such as EEG, PET and MEG. An emerging paradigm in neuroscience is that cognitive tasks are performed not by individual brain regions working in isolation but by networks consisting of several discrete brain regions that are said to be "functionally connected". Functional connectivity networks may be found using algorithms such as cluster analysis, spatial independent component analysis (ICA), seed based, and others. Synchronized brain regions may also be identified using long-range synchronization of the EEG, MEG, or other dynamic brain signals.<br />
<br />
'''<font color="#ff8000">大规模脑网络Large-scale brain networks</font>'''(也称为'''<font color="#ff8000">内在大脑网络Intrinsic brain networks</font>''')是在对基于'''<font color="#ff8000">血氧水平依赖效应BOLD</font>'''的'''<font color="#ff8000">功能性磁共振成像fMRI</font>'''信号<ref name="Riedl" />的统计分析或其他记录方法(如'''<font color="#ff8000">脑电图EEG<ref name=":1" /></font>'''、'''<font color="#ff8000">正电子发射断层扫描技术PET<ref name=":2" /></font>'''和'''<font color="#ff8000">脑磁图MEG<ref name=":3" /></font>''')中，表现出'''<font color="#ff8000">功能连接Functional connectivity</font>'''的'''<font color="#ff8000">脑区Brain regions</font>'''的集合。根据神经科学中一个新出现的范式，认知任务不是由单个脑区独立执行的，而是由几个互不相连的脑区“功能连接”组成的网络执行的。功能连接网络可以通过'''<font color="#ff8000">数据聚类Cluster analysis</font>'''、空间'''<font color="#ff8000">独立元素分析ICA</font>'''、种子点方法等算法来发现。<ref name="Petersen" />同步的脑区也可以用脑电图、脑磁图或其他动态脑信号的远程同步来识别。<ref name="Bressler" /><br />
<br />
The set of identified brain areas that are linked together in a large-scale network varies with cognitive function.<ref name="Bressler2">{{cite journal|last1=Bressler|first1=Steven L.|title=Neurocognitive networks|journal=Scholarpedia|volume=3|issue=2|pages=1567|doi=10.4249/scholarpedia.1567|year=2008|bibcode=2008SchpJ...3.1567B|doi-access=free}}</ref> When the cognitive state is not explicit (i.e., the subject is at "rest"), the large-scale brain network is a [[Resting state fMRI|resting state]] network (RSN). As a physical system with graph-like properties,<ref name="Bressler" /> a large-scale brain network has both nodes and edges and cannot be identified simply by the co-activation of brain areas. In recent decades, the analysis of brain networks was made feasible by advances in imaging techniques as well as new tools from [[graph theory]] and [[Dynamical systems theory|dynamical systems]].<br />
<br />
The set of identified brain areas that are linked together in a large-scale network varies with cognitive function. When the cognitive state is not explicit (i.e., the subject is at "rest"), the large-scale brain network is a resting state network (RSN). As a physical system with graph-like properties, a large-scale brain network has both nodes and edges and cannot be identified simply by the co-activation of brain areas. In recent decades, the analysis of brain networks was made feasible by advances in imaging techniques as well as new tools from graph theory and dynamical systems.<br />
<br />
大规模脑网络中,连接在一起的脑区集合因认知功能的不同而不同。<ref name="Bressler2" />当认知状态不明确(即主体处于“静止”状态)时，大规模脑网络是一个'''<font color="#ff8000">静息状态Resting state</font>'''网络(RSN)。作为一个具有图形特征的物理系统，<ref name="Bressler" />大规模脑网络既有节点又有边，不能简单地通过脑区的共同激活来识别。近几十年来，成像技术不断进步，同时'''<font color="#ff8000">图论Graph theory</font>'''、'''<font color="#ff8000">动力学系统Dynamical systems</font>'''领域出现了新的技术手段，这使得脑网络分析变得可行。<br />
<br />
Large-scale brain networks are identified by their function and provide a coherent framework for understanding [[cognition]] by offering a neural model of how different cognitive functions emerge when different sets of brain regions join together as self-organized coalitions. The number and composition of the coalitions will vary with the algorithm and parameters used to identify them.<ref name="Yeo" /><ref name=":4">{{cite journal|last1=Abou Elseoud|first1=Ahmed|last2=Littow|first2=Harri|last3=Remes|first3=Jukka|last4=Starck|first4=Tuomo|last5=Nikkinen|first5=Juha|last6=Nissilä|first6=Juuso|last7=Timonen|first7=Markku|last8=Tervonen|first8=Osmo|last9=Kiviniemi1|first9=Vesa|title=Group-ICA Model Order Highlights Patterns of Functional Brain Connectivity|journal= Frontiers in Systems Neuroscience|date=2011-06-03|volume=5|pages=37|doi=10.3389/fnsys.2011.00037|pmid=21687724|pmc=3109774 |doi-access=free}}</ref> In one model, there is only the [[default mode network]] and the [[task-positive network]], but most current analyses show several networks, from a small handful to 17.<ref name="Yeo" /> The most common and stable networks are enumerated below. The regions participating in a functional network may be dynamically reconfigured.<ref name="Petersen" /><ref name="Bassett">{{cite journal|last1=Bassett|first1=Daniella|last2=Bertolero|first2=Max|title=How Matter Becomes Mind|journal=Scientific American|date=July 2019|volume=321|issue=1|page=32|url=https://www.scientificamerican.com/|accessdate=23 June 2019 }}</ref><br />
<br />
Large-scale brain networks are identified by their function and provide a coherent framework for understanding cognition by offering a neural model of how different cognitive functions emerge when different sets of brain regions join together as self-organized coalitions. The number and composition of the coalitions will vary with the algorithm and parameters used to identify them. In one model, there is only the default mode network and the task-positive network, but most current analyses show several networks, from a small handful to 17. The most common and stable networks are enumerated below. The regions participating in a functional network may be dynamically reconfigured.<br />
<br />
大规模脑网络是通过其功能来进行识别的。通过研究大规模脑网络建立神经模型，对不同脑区组合所形成的自组织联合体如何实现不同的'''<font color="#ff8000">认知Cognition</font>'''功能进行解释，就能够为认知理解提供一个连贯的框架。识别算法和参数的不同会导致所识别出的上述联合体的数量和组成有所不同。<ref name="Yeo" /><ref name=":4" />一个模型理论认为，符合上述条件的神经模型只包含'''<font color="#ff8000">默认模式网络Default mode network</font>'''和'''<font color="#ff8000">任务激活网络Task-positive network</font>'''，但目前'''<font color="#32CD32">大多数分析理论都包括从几个到17个不等的网络。<ref name="Yeo" /></font>'''下面列举了最常见且稳定的网络。'''<font color="#32CD32"> 人脑</font>'''可以动态地重新配置参与功能网络的脑区。<ref name="Petersen" /><ref name="Bassett" /> <br />
<br />
Disruptions in activity in various networks have been implicated in neuropsychiatric disorders such as [[Depression (mood)|depression]], [[Alzheimer's disease|Alzheimer's]], [[Autism-spectrum disorder|autism spectrum disorder]], [[schizophrenia]], [[ADHD]]<ref name=":5">{{cite journal |last1=Griffiths |first1=Kristi R. |last2=Braund |first2=Taylor A. |last3=Kohn |first3=Michael R. |last4=Clarke |first4=Simon |last5=Williams |first5=Leanne M. |last6=Korgaonkar |first6=Mayuresh S. |title=Structural brain network topology underpinning ADHD and response to methylphenidate treatment |journal=Translational Psychiatry |date=2 March 2021 |volume=11 |issue=1 |pages=1–9 |doi=10.1038/s41398-021-01278-x | pmc=7925571 |pmid=33654073 |url=https://www.nature.com/articles/s41398-021-01278-x#citeas |access-date=16 November 2021}}</ref> and [[bipolar disorder]].<ref name=":6">{{Cite journal|url=https://www.researchgate.net/publication/51639686|title=Large-scale brain networks and psychopathology: A unifying triple network model|last=Menon|first=Vinod|s2cid=26653572|journal=Trends in Cognitive Sciences|date=2011-09-09|volume=15|issue=10|pages=483–506|doi=10.1016/j.tics.2011.08.003|pmid=21908230}}</ref><br />
<br />
Disruptions in activity in various networks have been implicated in neuropsychiatric disorders such as depression, Alzheimer's, autism spectrum disorder, schizophrenia, ADHD and bipolar disorder.<br />
<br />
脑网络活动的中断与诸多神经精神疾病密切相关，如'''<font color="#ff8000">抑郁症Depression</font>'''、'''<font color="#ff8000">老年痴呆症Alzheimer's</font>'''、'''<font color="#ff8000">自闭症谱系障碍Autism spectrum disorder</font>'''、'''<font color="#ff8000">精神分裂症Schizophrenia</font>'''、'''<font color="#ff8000">多动症ADHD<ref name=":5" /></font>'''和'''<font color="#ff8000">躁郁症Bipolar disorder<ref name=":6" /></font>'''。<br />
<br />
== Core networks==<br />
[[File:Heine2012x3010.png|thumb|An example that identified 10 large-scale brain networks from [[resting state fMRI]] activity through [[independent component analysis]].<ref name="Heine" />|链接=Special:FilePath/Heine2012x3010.png]]<br />
Because brain networks can be identified at various different resolutions and with various different neurobiological properties, there is no such thing as a universal atlas of brain networks that fits all circumstances.<ref>{{cite journal|last1=Eickhoff|first1=SB|last2=Yeo|first2=BTT|last3=Genon|first3=S|title=Imaging-based parcellations of the human brain.|journal=Nature Reviews. Neuroscience|date=November 2018|volume=19|issue=11|pages=672–686|doi=10.1038/s41583-018-0071-7|pmid=30305712|s2cid=52954265|url=http://juser.fz-juelich.de/record/856633/files/Eickhoff_Yeo_Genon_NRN_MainManuscriptInclFigures.pdf}}</ref> While acknowledging this problem, Uddin, Yeo, and Spreng proposed in 2019<ref name="Uddin2019">{{cite journal|last1=Uddin|first1=LQ|last2=Yeo|first2=BTT|last3=Spreng|first3=RN|title=Towards a Universal Taxonomy of Macro-scale Functional Human Brain Networks.|journal=Brain Topography|date=November 2019|volume=32|issue=6|pages=926–942|doi=10.1007/s10548-019-00744-6|pmid=31707621|pmc=7325607}}</ref> that the following six networks should be defined as core networks based on converging evidences from multiple studies<ref>{{cite journal|last1=Doucet|first1=GE|last2=Lee|first2=WH|last3=Frangou|first3=S|title=Evaluation of the spatial variability in the major resting-state networks across human brain functional atlases.|journal=Human Brain Mapping|date=2019-10-15|volume=40|issue=15|pages=4577–4587|doi=10.1002/hbm.24722|pmid=31322303|pmc=6771873}}</ref><ref name="Yeo" /><ref>{{cite journal|last1=Smith|first1=SM|last2=Fox|first2=PT|last3=Miller|first3=KL|last4=Glahn|first4=DC|last5=Fox|first5=PM|last6=Mackay|first6=CE|last7=Filippini|first7=N|last8=Watkins|first8=KE|last9=Toro|first9=R|last10=Laird|first10=AR|last11=Beckmann|first11=CF|title=Correspondence of the brain's functional architecture during activation and rest.|journal=Proceedings of the National Academy of Sciences of the United States of America|date=2009-08-04|volume=106|issue=31|pages=13040–5|doi=10.1073/pnas.0905267106|pmid=19620724|pmc=2722273|bibcode=2009PNAS..10613040S|doi-access=free}}</ref> to facilitate communication between researchers.<br />
<br />
<br />
Because brain networks can be identified at various different resolutions and with various different neurobiological properties, there is no such thing as a universal atlas of brain networks that fits all circumstances. While acknowledging this problem, Uddin, Yeo, and Spreng proposed in 2019 that the following six networks should be defined as core networks based on converging evidences from multiple studies to facilitate communication between researchers.<br />
<br />
因为大脑网络可以用不同的分辨率和不同的神经生物学特性来识别，所以没有适合所有情况的通用大脑网络图谱。在承认这个问题的同时，Uddin，Yeo，和 Spreng 在2019年提出，以下六个网络应该被定义为核心网络，基于来自多个研究的聚合证据，以促进研究人员之间的交流。<br />
<br />
===Default Mode (Medial frontoparietal)===<br />
{{Main|Default mode network}}<br />
*The default mode network is active when an individual is awake and at rest. It preferentially activates when individuals focus on internally-oriented tasks such as daydreaming, envisioning the future, retrieving memories, and [[theory of mind]]. It is negatively correlated with brain systems that focus on external visual signals. It is the most widely researched network.<ref name="Bressler" /><ref name="Bassett" /><ref>{{Cite journal|date=2012-08-15|title=The serendipitous discovery of the brain's default network|journal=NeuroImage|language=en|volume=62|issue=2|pages=1137–1145|doi=10.1016/j.neuroimage.2011.10.035|pmid=22037421|issn=1053-8119|last1=Buckner|first1=Randy L.|s2cid=9880586}}</ref><ref name="Riedl" /><ref name="Yuan"><br />
{{cite journal|last1=Yuan|first1=Rui|last2=Di|first2=Xin|last3=Taylor|first3=Paul A.|last4=Gohel|first4=Suril|last5=Tsai|first5=Yuan-Hsiung|last6=Biswal|first6=Bharat B.|title=Functional topography of the thalamocortical system in human|journal=Brain Structure and Function|date=30 April 2015|doi=10.1007/s00429-015-1018-7|pmid=25924563|pmc=6363530|volume=221|issue=4|pages=1971–1984}}</ref><ref name="Bell">{{cite journal|last1=Bell|first1=Peter T.|last2=Shine|first2=James M.|title=Estimating Large-Scale Network Convergence in the Human Functional Connectome|journal=Brain Connectivity|date=2015-11-09|volume=5|issue=9|doi=10.1089/brain.2015.0348|pmid=26005099|pages=565–74}}</ref><ref name="Heine">{{cite journal|last1=Heine|first1=Lizette|last2=Soddu|first2=Andrea|last3=Gomez|first3=Francisco|last4=Vanhaudenhuyse|first4=Audrey|last5=Tshibanda|first5=Luaba|last6=Thonnard|first6=Marie|last7=Charland-Verville|first7=Vanessa|last8=Kirsch|first8=Murielle|last9=Laureys|first9=Steven|last10=Demertzi|first10=Athena|title=Resting state networks and consciousness. Alterations of multiple resting state network connectivity in physiological, pharmacological and pathological consciousness states.|journal=Frontiers in Psychology|date=2012|volume=3|pages=295|doi=10.3389/fpsyg.2012.00295|pmid=22969735|pmc=3427917|doi-access=free}}</ref><ref name="Yeo">{{cite journal|last1=Yeo|first1=B. T. Thomas|last2=Krienen|first2=Fenna M.|last3=Sepulcre|first3=Jorge|last4=Sabuncu|first4=Mert R.|last5=Lashkari|first5=Danial|last6=Hollinshead|first6=Marisa|last7=Roffman|first7=Joshua L.|last8=Smoller|first8=Jordan W.|last9=Zöllei|first9=Lilla|last10=Polimeni|first10=Jonathan R.|last11=Fischl|first11=Bruce|last12=Liu|first12=Hesheng|last13=Buckner|first13=Randy L.|title=The organization of the human cerebral cortex estimated by intrinsic functional connectivity|journal=Journal of Neurophysiology|date=2011-09-01|volume=106|issue=3|pages=1125–1165|doi=10.1152/jn.00338.2011|pmid=21653723|pmc=3174820|bibcode=2011NatSD...2E0031H }}</ref><ref name="Shafiei">{{cite journal|last1=Shafiei|first1=Golia|last2=Zeighami|first2=Yashar|last3=Clark|first3=Crystal A.|last4=Coull|first4=Jennifer T.|last5=Nagano-Saito|first5=Atsuko|last6=Leyton|first6=Marco|last7=Dagher|first7=Alain|last8=Mišić|first8=Bratislav|title=Dopamine Signaling Modulates the Stability and Integration of Intrinsic Brain Networks|journal=Cerebral Cortex|date=2018-10-01|volume=29|issue=1|pages=397–409|doi=10.1093/cercor/bhy264|pmid=30357316|pmc=6294404 }}</ref><ref name="Bailey">{{cite journal|last1=Bailey|first1=Stephen K.|last2=Aboud|first2=Katherine S.|last3=Nguyen|first3=Tin Q.|last4=Cutting|first4=Laurie E.|title=Applying a network framework to the neurobiology of reading and dyslexia|journal=Journal of Neurodevelopmental Disorders|date=13 December 2018|volume=10|issue=1|page=37|doi=10.1186/s11689-018-9251-z|pmid=30541433|pmc=6291929 }}</ref><br />
<br />
<br />
*The default mode network is active when an individual is awake and at rest. It preferentially activates when individuals focus on internally-oriented tasks such as daydreaming, envisioning the future, retrieving memories, and theory of mind. It is negatively correlated with brain systems that focus on external visual signals. It is the most widely researched network.<br />
<br />
<br />
=<nowiki>= = 默认模式(Medial frontoparietal) = =</nowiki>= <br />
*默认模式网络在个人清醒和休息时处于活动状态。当个体专注于面向内部的任务时，比如做白日梦、展望未来、回忆和心理理论，它就会优先激活。它与专注于外部视觉信号的大脑系统负相关。它是研究最广泛的网络。<br />
<br />
===Salience (Midcingulo-Insular)===<br />
{{Main|Salience network}}<br />
*The salience network consists of several structures, including the anterior (bilateral) insula, dorsal anterior cingulate cortex, and three subcortical structures which are the ventral striatum, substantia nigra/ventral tegmental region.<ref>{{Cite journal|last1=Steimke|first1=Rosa|last2=Nomi|first2=Jason S.|last3=Calhoun|first3=Vince D.|last4=Stelzel|first4=Christine|last5=Paschke|first5=Lena M.|last6=Gaschler|first6=Robert|last7=Goschke|first7=Thomas|last8=Walter|first8=Henrik|last9=Uddin|first9=Lucina Q.|date=2017-12-01|title=Salience network dynamics underlying successful resistance of temptation|journal=Social Cognitive and Affective Neuroscience|language=en|volume=12|issue=12|pages=1928–1939|doi=10.1093/scan/nsx123|pmid=29048582|pmc=5716209|issn=1749-5016|doi-access=free}}</ref><ref name=":0">{{Citation|last=Menon|first=V.|title=Brain Mapping|chapter=Salience Network|date=2015-01-01|chapter-url=http://www.sciencedirect.com/science/article/pii/B978012397025100052X|pages=597–611|editor-last=Toga|editor-first=Arthur W.|publisher=Academic Press|isbn=978-0-12-397316-0|access-date=2019-12-08|doi=10.1016/B978-0-12-397025-1.00052-X}}</ref> It plays the key role of monitoring the [[Salience (neuroscience)|salience]] of external inputs and internal brain events.<ref name="Riedl" /><ref name="Bressler" /><ref name="Bassett" /><ref name="Yuan" /><ref name="Heine" /><ref name="Yeo" /><ref name="Shafiei" /> Specifically, it aids in directing attention by identifying important biological and cognitive events.<ref name=":0" /><ref name="Bailey" /><br />
*This network includes the ventral attention network, which primarily includes the [[temporoparietal junction]] and the ventral [[frontal cortex]] of the right hemisphere.<ref name="Uddin2019" /><ref name="Vossel" /> These areas respond when behaviorally relevant stimuli occur unexpectedly.<ref name="Vossel" /> The ventral attention network is inhibited during focused attention in which top-down processing is being used, such as when visually searching for something. This response may prevent goal-driven attention from being distracted by non-relevant stimuli. It becomes active again when the target or relevant information about the target is found.<ref name="Vossel" /><ref>{{Cite journal|last1=Shulman|first1=Gordon L.|last2=McAvoy|first2=Mark P.|last3=Cowan|first3=Melanie C.|last4=Astafiev|first4=Serguei V.|last5=Tansy|first5=Aaron P.|last6=d'Avossa|first6=Giovanni|last7=Corbetta|first7=Maurizio|date=2003-11-01|title=Quantitative Analysis of Attention and Detection Signals During Visual Search|journal=Journal of Neurophysiology|volume=90|issue=5|pages=3384–3397|doi=10.1152/jn.00343.2003|pmid=12917383|issn=0022-3077}}</ref><br />
<br />
<br />
*The salience network consists of several structures, including the anterior (bilateral) insula, dorsal anterior cingulate cortex, and three subcortical structures which are the ventral striatum, substantia nigra/ventral tegmental region. It plays the key role of monitoring the salience of external inputs and internal brain events. Specifically, it aids in directing attention by identifying important biological and cognitive events.<br />
*This network includes the ventral attention network, which primarily includes the temporoparietal junction and the ventral frontal cortex of the right hemisphere. These areas respond when behaviorally relevant stimuli occur unexpectedly. The ventral attention network is inhibited during focused attention in which top-down processing is being used, such as when visually searching for something. This response may prevent goal-driven attention from being distracted by non-relevant stimuli. It becomes active again when the target or relevant information about the target is found.<br />
<br />
突显网络由前(双)岛、前扣带回皮层和腹侧纹状体、黑质/腹侧被盖区3个皮层下结构组成。它起着监测外部输入和内部脑事件的显著性的关键作用。具体来说，它通过识别重要的生物学和认知事件来帮助引导注意力。<br />
*这个网络包括腹侧注意网络，主要包括右半球的颞顶联合区和腹侧额叶皮层。当行为相关的刺激意外发生时，这些区域会做出反应。腹侧注意网络在使用自上而下加工的集中注意过程中被抑制，例如在视觉搜索某物时。这种反应可以防止目标驱动的注意力被非相关的刺激分散。当找到目标或关于目标的相关信息时，它再次激活。<br />
<br />
===Attention (Dorsal frontoparietal) ===<br />
{{Main|Dorsal attention network}}<br />
*This network is involved in the voluntary, top-down deployment of attention.<ref name="Riedl" /><ref name="Yuan" /><ref name="Bell" /><ref name="Yeo" /><ref name="Shafiei" /><ref name="Vossel">{{cite journal|last1=Vossel|first1=Simone|last2=Geng|first2=Joy J.|last3=Fink|first3=Gereon R.|title=Dorsal and Ventral Attention Systems: Distinct Neural Circuits but Collaborative Roles|journal=The Neuroscientist|date=2014|volume=20|issue=2|pages=150–159|doi=10.1177/1073858413494269|pmid=23835449|pmc=4107817}}</ref><ref name="Hutton">{{cite journal|last1=Hutton|first1=John S.|last2=Dudley|first2=Jonathan|last3=Horowitz-Kraus|first3=Tzipi|last4=DeWitt|first4=Tom|last5=Holland|first5=Scott K.|title=Functional Connectivity of Attention, Visual, and Language Networks During Audio, Illustrated, and Animated Stories in Preschool-Age Children|journal=Brain Connectivity|date=1 September 2019|volume=9|issue=7|pages=580–592|doi=10.1089/brain.2019.0679|pmid=31144523|pmc=6775495|ref=Hutton}}</ref> Within the dorsal attention network, the intraparietal sulcus and frontal eye fields influence the visual areas of the brain. These influencing factors allow for the orientation of attention.<ref>{{Cite journal|last1=Fox|first1=Michael D.|last2=Corbetta|first2=Maurizio|last3=Snyder|first3=Abraham Z.|last4=Vincent|first4=Justin L.|last5=Raichle|first5=Marcus E.|date=2006-06-27|title=Spontaneous neuronal activity distinguishes human dorsal and ventral attention systems|journal=Proceedings of the National Academy of Sciences|language=en|volume=103|issue=26|pages=10046–10051|doi=10.1073/pnas.0604187103|issn=0027-8424|pmid=16788060|pmc=1480402|bibcode=2006PNAS..10310046F|doi-access=free}}</ref><ref name="Vossel" /><ref name="Bailey" /><br />
<br />
<br />
*This network is involved in the voluntary, top-down deployment of attention. Within the dorsal attention network, the intraparietal sulcus and frontal eye fields influence the visual areas of the brain. These influencing factors allow for the orientation of attention.<br />
<br />
=<nowiki>= = 注意力(背侧额顶骨) = =</nowiki>= <br />
*这个网络参与了自发的、自上而下的注意力分配。在背侧注意网络中，顶内沟和额眼影响大脑的视觉区域。这些影响因素决定了注意力的方向。<br />
<br />
===Control (Lateral frontoparietal)===<br />
{{Main|Frontoparietal network}}<br />
*This network initiates and modulates cognitive control and comprises 18 sub-regions of the brain.<ref>{{Cite journal|last1=Scolari|first1=Miranda|last2=Seidl-Rathkopf|first2=Katharina N|last3=Kastner|first3=Sabine|date=2015-02-01|title=Functions of the human frontoparietal attention network: Evidence from neuroimaging|journal=Current Opinion in Behavioral Sciences|series=Cognitive control|volume=1|pages=32–39|doi=10.1016/j.cobeha.2014.08.003|issn=2352-1546|pmid=27398396|pmc=4936532}}</ref> There is a strong correlation between fluid intelligence and the involvement of the fronto-parietal network with other networks.<ref>{{Cite journal|last1=Marek|first1=Scott|last2=Dosenbach|first2=Nico U. F.|date=June 2018|title=The frontoparietal network: function, electrophysiology, and importance of individual precision mapping|journal=Dialogues in Clinical Neuroscience|volume=20|issue=2|pages=133–140|doi=10.31887/DCNS.2018.20.2/smarek|issn=1294-8322|pmc=6136121|pmid=30250390}}</ref><br />
<br />
<br />
*This network initiates and modulates cognitive control and comprises 18 sub-regions of the brain. There is a strong correlation between fluid intelligence and the involvement of the fronto-parietal network with other networks.<br />
<br />
=<nowiki>= = 控制(侧额顶骨) = =</nowiki>= <br />
*这个网络启动和调节认知控制，包括大脑的18个亚区。在流体智力和额顶网络与其他网络的参与之间有很强的相关性。<br />
<br />
*Versions of this network have also been called the central executive (or executive control) network and the cognitive control network.<ref name="Uddin2019" /><br />
<br />
* Versions of this network have also been called the central executive (or executive control) network and the cognitive control network.<br />
<br />
<br />
*这种网络的版本也被称为中央执行(或执行控制)网络和认知控制网络。<br />
<br />
===Sensorimotor or Somatomotor (Pericentral)===<br />
{{Main|Sensorimotor network}}<br />
* This network processes somatosensory information and coordinates motion.<ref name="Heine" /><ref name="Yeo" /><ref name="Shafiei" /><ref name="Bassett" /><ref name="Yuan" /> The [[auditory cortex]] may be included.<ref name="Uddin2019" /><ref name="Yeo" /><br />
<br />
<br />
*This network processes somatosensory information and coordinates motion. The auditory cortex may be included.<br />
<br />
===Sensorimotor or Somatomotor (Pericentral)===<br />
<br />
<br />
*This network processes somatosensory information and coordinates motion.听觉皮层可能也包括在内。<br />
<br />
===Visual (Occipital)===<br />
{{See|Visual cortex}}<br />
*This network handles visual information processing.<ref name="Yang">{{cite journal|last1=Yang|first1=Yan-li|last2=Deng|first2=Hong-xia|last3=Xing|first3=Gui-yang|last4=Xia|first4=Xiao-luan|last5=Li|first5=Hai-fang|title=Brain functional network connectivity based on a visual task: visual information processing-related brain regions are significantly activated in the task state|journal=Neural Regeneration Research|date=2015|volume=10|issue=2|pages=298–307|doi=10.4103/1673-5374.152386|pmid=25883631|pmc=4392680 }}</ref><br />
<br />
<br />
*This network handles visual information processing.<br />
<br />
= = = = = = = = = = = = = 这个网络处理视觉信息。<br />
<br />
==Other networks==<br />
Different methods and data have identified several other brain networks, many of which greatly overlap or are subsets of more well-characterized core networks.<ref name="Uddin2019" /><br />
*Limbic<ref name="Bassett" /><ref name="Yeo" /><ref name="Bailey" /><br />
*Auditory<ref name="Yuan" /><ref name="Heine" /><br />
*Right/left executive<ref name="Yuan" /><ref name="Heine" /><br />
*Cerebellar<ref name="Bell" /><ref name="Heine" /><br />
*Spatial attention<ref name="Riedl" /><ref name="Bressler" /><br />
*Language<ref name="Bressler" /><ref name="Hutton" /><br />
*Lateral visual<ref name="Yuan" /><ref name="Bell" /><ref name="Heine" /><br />
*Temporal<ref name="Yeo" /><ref name="Shafiei" /><br />
*Visual perception/imagery<ref name="Hutton" /><br />
<br />
Different methods and data have identified several other brain networks, many of which greatly overlap or are subsets of more well-characterized core networks.<br />
*Limbic<br />
*Auditory<br />
*Right/left executive<br />
*Cerebellar<br />
*Spatial attention<br />
*Language<br />
*Lateral visual<br />
*Temporal<br />
*Visual perception/imagery<br />
<br />
其他网络不同的方法和数据已经确定了其他几个大脑网络，其中许多网络极大地重叠或者是更具特色的核心网络的子集。<br />
*边缘<br />
*听觉<br />
*右/左执行<br />
*小脑<br />
*空间注意<br />
*语言<br />
*外侧视觉<br />
*颞视知觉/图像<br />
<br />
==See also==<br />
*[[Complex network]]<br />
*[[Neural network]]<br />
<br />
*Complex network<br />
*Neural network<br />
<br />
=<nowiki>= = =</nowiki>= <br />
* 复杂网络<br />
*<br />
*神经网络<br />
*<br />
<br />
==References==<br />
{{reflist|30em}}<br />
<br />
{{Human connectomics}}<br />
<br />
[[Category:Neuroscience]]<br />
[[Category:Neural coding]]<br />
[[Category:Neural circuits]]<br />
[[Category:Neurophysiology]]<br />
<br />
Category:Neuroscience<br />
Category:Neural coding<br />
Category:Neural circuits<br />
Category:Neurophysiology<br />
<br />
类别: 神经科学类别: 神经编码类别: 神经回路类别: 神经生理学<br />
<br />
<noinclude><br />
<br />
<small>This page was moved from [[wikipedia:en:Large-scale brain network]]. Its edit history can be viewed at [[大规模脑网络/edithistory]]</small></noinclude><br />
<br />
[[Category:待整理页面]]</div>Shenky20https://wiki.swarma.org/index.php?title=%E5%A4%A7%E8%A7%84%E6%A8%A1%E8%84%91%E7%BD%91%E7%BB%9C&diff=29985大规模脑网络2022-04-04T07:55:36Z<p>Shenky20：</p>
<hr />
<div>此词条由神经动力学模型读书会词条梳理志愿者Shenky20翻译审校，翻译字数共800，未经专家审核，带来阅读不便，请见谅。<br />
<br />
{{Short description|Collections of brain regions}}<br />
'''Large-scale brain networks''' (also known as '''intrinsic brain networks''') are collections of widespread [[brain regions]] showing [[Resting state fMRI#Functional|functional connectivity]] by statistical analysis of the [[Functional magnetic resonance imaging|fMRI]] [[BOLD signal]]<ref name=Riedl>{{cite journal|last1=Riedl|first1=Valentin|last2=Utz|first2=Lukas|last3=Castrillón|first3=Gabriel|last4=Grimmer|first4=Timo|last5=Rauschecker|first5=Josef P.|last6=Ploner|first6=Markus|last7=Friston|first7=Karl J.|last8=Drzezga|first8=Alexander|last9=Sorg|first9=Christian|title=Metabolic connectivity mapping reveals effective connectivity in the resting human brain|journal=PNAS|date=January 12, 2016|volume=113|issue=2|pages=428–433|doi=10.1073/pnas.1513752113|pmid=26712010|pmc=4720331|bibcode=2016PNAS..113..428R|doi-access=free}}</ref> or other recording methods such as [[Electroencephalography|EEG]],<ref>{{Cite journal|last1=Foster|first1=Brett L.|last2=Parvizi|first2=Josef|date=2012-03-01|title=Resting oscillations and cross-frequency coupling in the human posteromedial cortex|journal=NeuroImage|volume=60|issue=1|pages=384–391|doi=10.1016/j.neuroimage.2011.12.019|pmid=22227048|issn=1053-8119|pmc=3596417}}</ref> [[Positron emission tomography|PET]]<ref>{{Cite journal|last1=Buckner|first1=Randy L.|last2=Andrews‐Hanna|first2=Jessica R.|last3=Schacter|first3=Daniel L.|date=2008|title=The Brain's Default Network|journal=Annals of the New York Academy of Sciences|language=en|volume=1124|issue=1|pages=1–38|doi=10.1196/annals.1440.011|pmid=18400922|issn=1749-6632|bibcode=2008NYASA1124....1B|s2cid=3167595}}</ref> and [[Magnetoencephalography|MEG]].<ref>{{Cite journal|last1=Morris|first1=Peter G.|last2=Smith|first2=Stephen M.|last3=Barnes|first3=Gareth R.|last4=Stephenson|first4=Mary C.|last5=Hale|first5=Joanne R.|last6=Price|first6=Darren|last7=Luckhoo|first7=Henry|last8=Woolrich|first8=Mark|last9=Brookes|first9=Matthew J.|date=2011-10-04|title=Investigating the electrophysiological basis of resting state networks using magnetoencephalography|journal=Proceedings of the National Academy of Sciences|language=en|volume=108|issue=40|pages=16783–16788|doi=10.1073/pnas.1112685108|issn=0027-8424|pmid=21930901|pmc=3189080|bibcode=2011PNAS..10816783B|doi-access=free}}</ref> An emerging paradigm in neuroscience is that cognitive tasks are performed not by individual brain regions working in isolation but by networks consisting of several discrete brain regions that are said to be "functionally connected". Functional connectivity networks may be found using algorithms such as [[cluster analysis]], spatial [[independent component analysis]] (ICA), seed based, and others.<ref name="Petersen">{{cite journal|last1=Petersen|first1=Steven|last2=Sporns|first2=Olaf|title=Brain Networks and Cognitive Architectures|journal=Neuron|date=October 2015|volume=88|issue=1|pages=207–219|doi=10.1016/j.neuron.2015.09.027|pmid=26447582|pmc=4598639 }}</ref> Synchronized brain regions may also be identified using long-range synchronization of the EEG, MEG, or other dynamic brain signals.<ref name=Bressler>{{cite journal|last1=Bressler|first1=Steven L.|last2=Menon|first2=Vinod|s2cid=5967761|title=Large scale brain networks in cognition: emerging methods and principles|journal=Trends in Cognitive Sciences|date=June 2010|volume=14|issue=6|pages=233–290|doi=10.1016/j.tics.2010.04.004|url=http://www.cell.com/trends/cognitive-sciences/issue?pii=S1364-6613(10)X0005-5|accessdate=24 January 2016|pmid=20493761}}</ref><br />
<br />
<br />
Large-scale brain networks (also known as intrinsic brain networks) are collections of widespread brain regions showing functional connectivity by statistical analysis of the fMRI BOLD signal or other recording methods such as EEG, PET and MEG. An emerging paradigm in neuroscience is that cognitive tasks are performed not by individual brain regions working in isolation but by networks consisting of several discrete brain regions that are said to be "functionally connected". Functional connectivity networks may be found using algorithms such as cluster analysis, spatial independent component analysis (ICA), seed based, and others. Synchronized brain regions may also be identified using long-range synchronization of the EEG, MEG, or other dynamic brain signals.<br />
<br />
'''<font color="#ff8000">大规模脑网络Large-scale brain networks</font>'''(也称为'''<font color="#ff8000">内在大脑网络Intrinsic brain networks</font>''')是在对基于'''<font color="#ff8000">血氧水平依赖效应BOLD</font>'''的'''<font color="#ff8000">功能性磁共振成像fMRI</font>'''信号的统计分析或其他记录方法(如'''<font color="#ff8000">脑电图EEG</font>'''、'''<font color="#ff8000">正电子发射断层扫描技术PET</font>'''和'''<font color="#ff8000">脑磁图MEG</font>''')中，表现出'''<font color="#ff8000">功能连接Functional connectivity</font>'''的'''<font color="#ff8000">脑区Brain regions</font>'''的集合。根据神经科学中一个新出现的范式，认知任务不是由单个脑区独立执行的，而是由几个互不相连的脑区“功能连接”组成的网络执行的。功能连接网络可以通过'''<font color="#ff8000">数据聚类Cluster analysis</font>'''、空间'''<font color="#ff8000">独立元素分析ICA</font>'''、种子点方法等算法来发现。同步的脑区也可以用脑电图、脑磁图或其他动态脑信号的远程同步来识别。<br />
<br />
The set of identified brain areas that are linked together in a large-scale network varies with cognitive function.<ref name="Bressler2">{{cite journal|last1=Bressler|first1=Steven L.|title=Neurocognitive networks|journal=Scholarpedia|volume=3|issue=2|pages=1567|doi=10.4249/scholarpedia.1567|year=2008|bibcode=2008SchpJ...3.1567B|doi-access=free}}</ref> When the cognitive state is not explicit (i.e., the subject is at "rest"), the large-scale brain network is a [[Resting state fMRI|resting state]] network (RSN). As a physical system with graph-like properties,<ref name="Bressler" /> a large-scale brain network has both nodes and edges and cannot be identified simply by the co-activation of brain areas. In recent decades, the analysis of brain networks was made feasible by advances in imaging techniques as well as new tools from [[graph theory]] and [[Dynamical systems theory|dynamical systems]].<br />
<br />
The set of identified brain areas that are linked together in a large-scale network varies with cognitive function. When the cognitive state is not explicit (i.e., the subject is at "rest"), the large-scale brain network is a resting state network (RSN). As a physical system with graph-like properties, a large-scale brain network has both nodes and edges and cannot be identified simply by the co-activation of brain areas. In recent decades, the analysis of brain networks was made feasible by advances in imaging techniques as well as new tools from graph theory and dynamical systems.<br />
<br />
大规模脑网络中,连接在一起的脑区集合因认知功能的不同而不同。当认知状态不明确(即主体处于“静止”状态)时，大规模脑网络是一个'''<font color="#ff8000">静息状态Resting state</font>'''网络(RSN)。作为一个具有图形特征的物理系统，大规模脑网络既有节点又有边，不能简单地通过脑区的共同激活来识别。近几十年来，成像技术不断进步，同时'''<font color="#ff8000">图论Graph theory</font>'''、'''<font color="#ff8000">动力学系统Dynamical systems</font>'''领域出现了新的技术手段，这使得脑网络分析变得可行。<br />
<br />
Large-scale brain networks are identified by their function and provide a coherent framework for understanding [[cognition]] by offering a neural model of how different cognitive functions emerge when different sets of brain regions join together as self-organized coalitions. The number and composition of the coalitions will vary with the algorithm and parameters used to identify them.<ref name="Yeo" /><ref>{{cite journal|last1=Abou Elseoud|first1=Ahmed|last2=Littow|first2=Harri|last3=Remes|first3=Jukka|last4=Starck|first4=Tuomo|last5=Nikkinen|first5=Juha|last6=Nissilä|first6=Juuso|last7=Timonen|first7=Markku|last8=Tervonen|first8=Osmo|last9=Kiviniemi1|first9=Vesa|title=Group-ICA Model Order Highlights Patterns of Functional Brain Connectivity|journal= Frontiers in Systems Neuroscience|date=2011-06-03|volume=5|pages=37|doi=10.3389/fnsys.2011.00037|pmid=21687724|pmc=3109774 |doi-access=free}}</ref> In one model, there is only the [[default mode network]] and the [[task-positive network]], but most current analyses show several networks, from a small handful to 17.<ref name="Yeo" /> The most common and stable networks are enumerated below. The regions participating in a functional network may be dynamically reconfigured.<ref name="Petersen" /><ref name="Bassett">{{cite journal|last1=Bassett|first1=Daniella|last2=Bertolero|first2=Max|title=How Matter Becomes Mind|journal=Scientific American|date=July 2019|volume=321|issue=1|page=32|url=https://www.scientificamerican.com/|accessdate=23 June 2019 }}</ref><br />
<br />
Large-scale brain networks are identified by their function and provide a coherent framework for understanding cognition by offering a neural model of how different cognitive functions emerge when different sets of brain regions join together as self-organized coalitions. The number and composition of the coalitions will vary with the algorithm and parameters used to identify them. In one model, there is only the default mode network and the task-positive network, but most current analyses show several networks, from a small handful to 17. The most common and stable networks are enumerated below. The regions participating in a functional network may be dynamically reconfigured.<br />
<br />
大规模脑网络是通过其功能来进行识别的。通过研究大规模脑网络建立神经模型，对不同脑区组合所形成的自组织联合体如何实现不同的'''<font color="#ff8000">认知Cognition</font>'''功能进行解释，就能够为认知理解提供一个连贯的框架。识别算法和参数的不同会导致所识别出的上述联合体的数量和组成有所不同。一个模型理论认为，符合上述条件的神经模型只包含'''<font color="#ff8000">默认模式网络Default mode network</font>'''和'''<font color="#ff8000">任务激活网络Task-positive network</font>'''，但目前'''<font color="#32CD32">大多数分析理论都包括从几个到17个不等的网络。</font>'''下面列举了最常见且稳定的网络。'''<font color="#32CD32"> 人脑</font>'''可以动态地重新配置参与功能网络的脑区。 <br />
<br />
Disruptions in activity in various networks have been implicated in neuropsychiatric disorders such as [[Depression (mood)|depression]], [[Alzheimer's disease|Alzheimer's]], [[Autism-spectrum disorder|autism spectrum disorder]], [[schizophrenia]], [[ADHD]]<ref>{{cite journal |last1=Griffiths |first1=Kristi R. |last2=Braund |first2=Taylor A. |last3=Kohn |first3=Michael R. |last4=Clarke |first4=Simon |last5=Williams |first5=Leanne M. |last6=Korgaonkar |first6=Mayuresh S. |title=Structural brain network topology underpinning ADHD and response to methylphenidate treatment |journal=Translational Psychiatry |date=2 March 2021 |volume=11 |issue=1 |pages=1–9 |doi=10.1038/s41398-021-01278-x | pmc=7925571 |pmid=33654073 |url=https://www.nature.com/articles/s41398-021-01278-x#citeas |access-date=16 November 2021}}</ref> and [[bipolar disorder]].<ref>{{Cite journal|url=https://www.researchgate.net/publication/51639686|title=Large-scale brain networks and psychopathology: A unifying triple network model|last=Menon|first=Vinod|s2cid=26653572|journal=Trends in Cognitive Sciences|date=2011-09-09|volume=15|issue=10|pages=483–506|doi=10.1016/j.tics.2011.08.003|pmid=21908230}}</ref><br />
<br />
Disruptions in activity in various networks have been implicated in neuropsychiatric disorders such as depression, Alzheimer's, autism spectrum disorder, schizophrenia, ADHD and bipolar disorder.<br />
<br />
脑网络活动的中断与诸多神经精神疾病密切相关，如'''<font color="#ff8000">抑郁症Depression</font>'''、'''<font color="#ff8000">老年痴呆症Alzheimer's</font>'''、'''<font color="#ff8000">自闭症谱系障碍Autism spectrum disorder</font>'''、'''<font color="#ff8000">精神分裂症Schizophrenia</font>'''、'''<font color="#ff8000">多动症ADHD</font>'''和'''<font color="#ff8000">躁郁症Bipolar disorder</font>'''。<br />
<br />
== Core networks==<br />
[[File:Heine2012x3010.png|thumb|An example that identified 10 large-scale brain networks from [[resting state fMRI]] activity through [[independent component analysis]].<ref name="Heine" />|链接=Special:FilePath/Heine2012x3010.png]]<br />
Because brain networks can be identified at various different resolutions and with various different neurobiological properties, there is no such thing as a universal atlas of brain networks that fits all circumstances.<ref>{{cite journal|last1=Eickhoff|first1=SB|last2=Yeo|first2=BTT|last3=Genon|first3=S|title=Imaging-based parcellations of the human brain.|journal=Nature Reviews. Neuroscience|date=November 2018|volume=19|issue=11|pages=672–686|doi=10.1038/s41583-018-0071-7|pmid=30305712|s2cid=52954265|url=http://juser.fz-juelich.de/record/856633/files/Eickhoff_Yeo_Genon_NRN_MainManuscriptInclFigures.pdf}}</ref> While acknowledging this problem, Uddin, Yeo, and Spreng proposed in 2019<ref name="Uddin2019">{{cite journal|last1=Uddin|first1=LQ|last2=Yeo|first2=BTT|last3=Spreng|first3=RN|title=Towards a Universal Taxonomy of Macro-scale Functional Human Brain Networks.|journal=Brain Topography|date=November 2019|volume=32|issue=6|pages=926–942|doi=10.1007/s10548-019-00744-6|pmid=31707621|pmc=7325607}}</ref> that the following six networks should be defined as core networks based on converging evidences from multiple studies<ref>{{cite journal|last1=Doucet|first1=GE|last2=Lee|first2=WH|last3=Frangou|first3=S|title=Evaluation of the spatial variability in the major resting-state networks across human brain functional atlases.|journal=Human Brain Mapping|date=2019-10-15|volume=40|issue=15|pages=4577–4587|doi=10.1002/hbm.24722|pmid=31322303|pmc=6771873}}</ref><ref name="Yeo" /><ref>{{cite journal|last1=Smith|first1=SM|last2=Fox|first2=PT|last3=Miller|first3=KL|last4=Glahn|first4=DC|last5=Fox|first5=PM|last6=Mackay|first6=CE|last7=Filippini|first7=N|last8=Watkins|first8=KE|last9=Toro|first9=R|last10=Laird|first10=AR|last11=Beckmann|first11=CF|title=Correspondence of the brain's functional architecture during activation and rest.|journal=Proceedings of the National Academy of Sciences of the United States of America|date=2009-08-04|volume=106|issue=31|pages=13040–5|doi=10.1073/pnas.0905267106|pmid=19620724|pmc=2722273|bibcode=2009PNAS..10613040S|doi-access=free}}</ref> to facilitate communication between researchers.<br />
<br />
<br />
Because brain networks can be identified at various different resolutions and with various different neurobiological properties, there is no such thing as a universal atlas of brain networks that fits all circumstances. While acknowledging this problem, Uddin, Yeo, and Spreng proposed in 2019 that the following six networks should be defined as core networks based on converging evidences from multiple studies to facilitate communication between researchers.<br />
<br />
因为大脑网络可以用不同的分辨率和不同的神经生物学特性来识别，所以没有适合所有情况的通用大脑网络图谱。在承认这个问题的同时，Uddin，Yeo，和 Spreng 在2019年提出，以下六个网络应该被定义为核心网络，基于来自多个研究的聚合证据，以促进研究人员之间的交流。<br />
<br />
===Default Mode (Medial frontoparietal)===<br />
{{Main|Default mode network}}<br />
*The default mode network is active when an individual is awake and at rest. It preferentially activates when individuals focus on internally-oriented tasks such as daydreaming, envisioning the future, retrieving memories, and [[theory of mind]]. It is negatively correlated with brain systems that focus on external visual signals. It is the most widely researched network.<ref name="Bressler" /><ref name="Bassett" /><ref>{{Cite journal|date=2012-08-15|title=The serendipitous discovery of the brain's default network|journal=NeuroImage|language=en|volume=62|issue=2|pages=1137–1145|doi=10.1016/j.neuroimage.2011.10.035|pmid=22037421|issn=1053-8119|last1=Buckner|first1=Randy L.|s2cid=9880586}}</ref><ref name="Riedl" /><ref name="Yuan"><br />
{{cite journal|last1=Yuan|first1=Rui|last2=Di|first2=Xin|last3=Taylor|first3=Paul A.|last4=Gohel|first4=Suril|last5=Tsai|first5=Yuan-Hsiung|last6=Biswal|first6=Bharat B.|title=Functional topography of the thalamocortical system in human|journal=Brain Structure and Function|date=30 April 2015|doi=10.1007/s00429-015-1018-7|pmid=25924563|pmc=6363530|volume=221|issue=4|pages=1971–1984}}</ref><ref name="Bell">{{cite journal|last1=Bell|first1=Peter T.|last2=Shine|first2=James M.|title=Estimating Large-Scale Network Convergence in the Human Functional Connectome|journal=Brain Connectivity|date=2015-11-09|volume=5|issue=9|doi=10.1089/brain.2015.0348|pmid=26005099|pages=565–74}}</ref><ref name="Heine">{{cite journal|last1=Heine|first1=Lizette|last2=Soddu|first2=Andrea|last3=Gomez|first3=Francisco|last4=Vanhaudenhuyse|first4=Audrey|last5=Tshibanda|first5=Luaba|last6=Thonnard|first6=Marie|last7=Charland-Verville|first7=Vanessa|last8=Kirsch|first8=Murielle|last9=Laureys|first9=Steven|last10=Demertzi|first10=Athena|title=Resting state networks and consciousness. Alterations of multiple resting state network connectivity in physiological, pharmacological and pathological consciousness states.|journal=Frontiers in Psychology|date=2012|volume=3|pages=295|doi=10.3389/fpsyg.2012.00295|pmid=22969735|pmc=3427917|doi-access=free}}</ref><ref name="Yeo">{{cite journal|last1=Yeo|first1=B. T. Thomas|last2=Krienen|first2=Fenna M.|last3=Sepulcre|first3=Jorge|last4=Sabuncu|first4=Mert R.|last5=Lashkari|first5=Danial|last6=Hollinshead|first6=Marisa|last7=Roffman|first7=Joshua L.|last8=Smoller|first8=Jordan W.|last9=Zöllei|first9=Lilla|last10=Polimeni|first10=Jonathan R.|last11=Fischl|first11=Bruce|last12=Liu|first12=Hesheng|last13=Buckner|first13=Randy L.|title=The organization of the human cerebral cortex estimated by intrinsic functional connectivity|journal=Journal of Neurophysiology|date=2011-09-01|volume=106|issue=3|pages=1125–1165|doi=10.1152/jn.00338.2011|pmid=21653723|pmc=3174820|bibcode=2011NatSD...2E0031H }}</ref><ref name="Shafiei">{{cite journal|last1=Shafiei|first1=Golia|last2=Zeighami|first2=Yashar|last3=Clark|first3=Crystal A.|last4=Coull|first4=Jennifer T.|last5=Nagano-Saito|first5=Atsuko|last6=Leyton|first6=Marco|last7=Dagher|first7=Alain|last8=Mišić|first8=Bratislav|title=Dopamine Signaling Modulates the Stability and Integration of Intrinsic Brain Networks|journal=Cerebral Cortex|date=2018-10-01|volume=29|issue=1|pages=397–409|doi=10.1093/cercor/bhy264|pmid=30357316|pmc=6294404 }}</ref><ref name="Bailey">{{cite journal|last1=Bailey|first1=Stephen K.|last2=Aboud|first2=Katherine S.|last3=Nguyen|first3=Tin Q.|last4=Cutting|first4=Laurie E.|title=Applying a network framework to the neurobiology of reading and dyslexia|journal=Journal of Neurodevelopmental Disorders|date=13 December 2018|volume=10|issue=1|page=37|doi=10.1186/s11689-018-9251-z|pmid=30541433|pmc=6291929 }}</ref><br />
<br />
<br />
*The default mode network is active when an individual is awake and at rest. It preferentially activates when individuals focus on internally-oriented tasks such as daydreaming, envisioning the future, retrieving memories, and theory of mind. It is negatively correlated with brain systems that focus on external visual signals. It is the most widely researched network.<br />
<br />
<br />
=<nowiki>= = 默认模式(Medial frontoparietal) = =</nowiki>= <br />
*默认模式网络在个人清醒和休息时处于活动状态。当个体专注于面向内部的任务时，比如做白日梦、展望未来、回忆和心理理论，它就会优先激活。它与专注于外部视觉信号的大脑系统负相关。它是研究最广泛的网络。<br />
<br />
===Salience (Midcingulo-Insular)===<br />
{{Main|Salience network}}<br />
*The salience network consists of several structures, including the anterior (bilateral) insula, dorsal anterior cingulate cortex, and three subcortical structures which are the ventral striatum, substantia nigra/ventral tegmental region.<ref>{{Cite journal|last1=Steimke|first1=Rosa|last2=Nomi|first2=Jason S.|last3=Calhoun|first3=Vince D.|last4=Stelzel|first4=Christine|last5=Paschke|first5=Lena M.|last6=Gaschler|first6=Robert|last7=Goschke|first7=Thomas|last8=Walter|first8=Henrik|last9=Uddin|first9=Lucina Q.|date=2017-12-01|title=Salience network dynamics underlying successful resistance of temptation|journal=Social Cognitive and Affective Neuroscience|language=en|volume=12|issue=12|pages=1928–1939|doi=10.1093/scan/nsx123|pmid=29048582|pmc=5716209|issn=1749-5016|doi-access=free}}</ref><ref name=":0">{{Citation|last=Menon|first=V.|title=Brain Mapping|chapter=Salience Network|date=2015-01-01|chapter-url=http://www.sciencedirect.com/science/article/pii/B978012397025100052X|pages=597–611|editor-last=Toga|editor-first=Arthur W.|publisher=Academic Press|isbn=978-0-12-397316-0|access-date=2019-12-08|doi=10.1016/B978-0-12-397025-1.00052-X}}</ref> It plays the key role of monitoring the [[Salience (neuroscience)|salience]] of external inputs and internal brain events.<ref name="Riedl" /><ref name="Bressler" /><ref name="Bassett" /><ref name="Yuan" /><ref name="Heine" /><ref name="Yeo" /><ref name="Shafiei" /> Specifically, it aids in directing attention by identifying important biological and cognitive events.<ref name=":0" /><ref name="Bailey" /><br />
*This network includes the ventral attention network, which primarily includes the [[temporoparietal junction]] and the ventral [[frontal cortex]] of the right hemisphere.<ref name="Uddin2019" /><ref name="Vossel" /> These areas respond when behaviorally relevant stimuli occur unexpectedly.<ref name="Vossel" /> The ventral attention network is inhibited during focused attention in which top-down processing is being used, such as when visually searching for something. This response may prevent goal-driven attention from being distracted by non-relevant stimuli. It becomes active again when the target or relevant information about the target is found.<ref name="Vossel" /><ref>{{Cite journal|last1=Shulman|first1=Gordon L.|last2=McAvoy|first2=Mark P.|last3=Cowan|first3=Melanie C.|last4=Astafiev|first4=Serguei V.|last5=Tansy|first5=Aaron P.|last6=d'Avossa|first6=Giovanni|last7=Corbetta|first7=Maurizio|date=2003-11-01|title=Quantitative Analysis of Attention and Detection Signals During Visual Search|journal=Journal of Neurophysiology|volume=90|issue=5|pages=3384–3397|doi=10.1152/jn.00343.2003|pmid=12917383|issn=0022-3077}}</ref><br />
<br />
<br />
*The salience network consists of several structures, including the anterior (bilateral) insula, dorsal anterior cingulate cortex, and three subcortical structures which are the ventral striatum, substantia nigra/ventral tegmental region. It plays the key role of monitoring the salience of external inputs and internal brain events. Specifically, it aids in directing attention by identifying important biological and cognitive events.<br />
*This network includes the ventral attention network, which primarily includes the temporoparietal junction and the ventral frontal cortex of the right hemisphere. These areas respond when behaviorally relevant stimuli occur unexpectedly. The ventral attention network is inhibited during focused attention in which top-down processing is being used, such as when visually searching for something. This response may prevent goal-driven attention from being distracted by non-relevant stimuli. It becomes active again when the target or relevant information about the target is found.<br />
<br />
突显网络由前(双)岛、前扣带回皮层和腹侧纹状体、黑质/腹侧被盖区3个皮层下结构组成。它起着监测外部输入和内部脑事件的显著性的关键作用。具体来说，它通过识别重要的生物学和认知事件来帮助引导注意力。<br />
*这个网络包括腹侧注意网络，主要包括右半球的颞顶联合区和腹侧额叶皮层。当行为相关的刺激意外发生时，这些区域会做出反应。腹侧注意网络在使用自上而下加工的集中注意过程中被抑制，例如在视觉搜索某物时。这种反应可以防止目标驱动的注意力被非相关的刺激分散。当找到目标或关于目标的相关信息时，它再次激活。<br />
<br />
===Attention (Dorsal frontoparietal) ===<br />
{{Main|Dorsal attention network}}<br />
*This network is involved in the voluntary, top-down deployment of attention.<ref name="Riedl" /><ref name="Yuan" /><ref name="Bell" /><ref name="Yeo" /><ref name="Shafiei" /><ref name="Vossel">{{cite journal|last1=Vossel|first1=Simone|last2=Geng|first2=Joy J.|last3=Fink|first3=Gereon R.|title=Dorsal and Ventral Attention Systems: Distinct Neural Circuits but Collaborative Roles|journal=The Neuroscientist|date=2014|volume=20|issue=2|pages=150–159|doi=10.1177/1073858413494269|pmid=23835449|pmc=4107817}}</ref><ref name="Hutton">{{cite journal|last1=Hutton|first1=John S.|last2=Dudley|first2=Jonathan|last3=Horowitz-Kraus|first3=Tzipi|last4=DeWitt|first4=Tom|last5=Holland|first5=Scott K.|title=Functional Connectivity of Attention, Visual, and Language Networks During Audio, Illustrated, and Animated Stories in Preschool-Age Children|journal=Brain Connectivity|date=1 September 2019|volume=9|issue=7|pages=580–592|doi=10.1089/brain.2019.0679|pmid=31144523|pmc=6775495|ref=Hutton}}</ref> Within the dorsal attention network, the intraparietal sulcus and frontal eye fields influence the visual areas of the brain. These influencing factors allow for the orientation of attention.<ref>{{Cite journal|last1=Fox|first1=Michael D.|last2=Corbetta|first2=Maurizio|last3=Snyder|first3=Abraham Z.|last4=Vincent|first4=Justin L.|last5=Raichle|first5=Marcus E.|date=2006-06-27|title=Spontaneous neuronal activity distinguishes human dorsal and ventral attention systems|journal=Proceedings of the National Academy of Sciences|language=en|volume=103|issue=26|pages=10046–10051|doi=10.1073/pnas.0604187103|issn=0027-8424|pmid=16788060|pmc=1480402|bibcode=2006PNAS..10310046F|doi-access=free}}</ref><ref name="Vossel" /><ref name="Bailey" /><br />
<br />
<br />
*This network is involved in the voluntary, top-down deployment of attention. Within the dorsal attention network, the intraparietal sulcus and frontal eye fields influence the visual areas of the brain. These influencing factors allow for the orientation of attention.<br />
<br />
=<nowiki>= = 注意力(背侧额顶骨) = =</nowiki>= <br />
*这个网络参与了自发的、自上而下的注意力分配。在背侧注意网络中，顶内沟和额眼影响大脑的视觉区域。这些影响因素决定了注意力的方向。<br />
<br />
===Control (Lateral frontoparietal)===<br />
{{Main|Frontoparietal network}}<br />
*This network initiates and modulates cognitive control and comprises 18 sub-regions of the brain.<ref>{{Cite journal|last1=Scolari|first1=Miranda|last2=Seidl-Rathkopf|first2=Katharina N|last3=Kastner|first3=Sabine|date=2015-02-01|title=Functions of the human frontoparietal attention network: Evidence from neuroimaging|journal=Current Opinion in Behavioral Sciences|series=Cognitive control|volume=1|pages=32–39|doi=10.1016/j.cobeha.2014.08.003|issn=2352-1546|pmid=27398396|pmc=4936532}}</ref> There is a strong correlation between fluid intelligence and the involvement of the fronto-parietal network with other networks.<ref>{{Cite journal|last1=Marek|first1=Scott|last2=Dosenbach|first2=Nico U. F.|date=June 2018|title=The frontoparietal network: function, electrophysiology, and importance of individual precision mapping|journal=Dialogues in Clinical Neuroscience|volume=20|issue=2|pages=133–140|doi=10.31887/DCNS.2018.20.2/smarek|issn=1294-8322|pmc=6136121|pmid=30250390}}</ref><br />
<br />
<br />
*This network initiates and modulates cognitive control and comprises 18 sub-regions of the brain. There is a strong correlation between fluid intelligence and the involvement of the fronto-parietal network with other networks.<br />
<br />
=<nowiki>= = 控制(侧额顶骨) = =</nowiki>= <br />
*这个网络启动和调节认知控制，包括大脑的18个亚区。在流体智力和额顶网络与其他网络的参与之间有很强的相关性。<br />
<br />
*Versions of this network have also been called the central executive (or executive control) network and the cognitive control network.<ref name="Uddin2019" /><br />
<br />
* Versions of this network have also been called the central executive (or executive control) network and the cognitive control network.<br />
<br />
<br />
*这种网络的版本也被称为中央执行(或执行控制)网络和认知控制网络。<br />
<br />
===Sensorimotor or Somatomotor (Pericentral)===<br />
{{Main|Sensorimotor network}}<br />
* This network processes somatosensory information and coordinates motion.<ref name="Heine" /><ref name="Yeo" /><ref name="Shafiei" /><ref name="Bassett" /><ref name="Yuan" /> The [[auditory cortex]] may be included.<ref name="Uddin2019" /><ref name="Yeo" /><br />
<br />
<br />
*This network processes somatosensory information and coordinates motion. The auditory cortex may be included.<br />
<br />
===Sensorimotor or Somatomotor (Pericentral)===<br />
<br />
<br />
*This network processes somatosensory information and coordinates motion.听觉皮层可能也包括在内。<br />
<br />
===Visual (Occipital)===<br />
{{See|Visual cortex}}<br />
*This network handles visual information processing.<ref name="Yang">{{cite journal|last1=Yang|first1=Yan-li|last2=Deng|first2=Hong-xia|last3=Xing|first3=Gui-yang|last4=Xia|first4=Xiao-luan|last5=Li|first5=Hai-fang|title=Brain functional network connectivity based on a visual task: visual information processing-related brain regions are significantly activated in the task state|journal=Neural Regeneration Research|date=2015|volume=10|issue=2|pages=298–307|doi=10.4103/1673-5374.152386|pmid=25883631|pmc=4392680 }}</ref><br />
<br />
<br />
*This network handles visual information processing.<br />
<br />
= = = = = = = = = = = = = 这个网络处理视觉信息。<br />
<br />
==Other networks==<br />
Different methods and data have identified several other brain networks, many of which greatly overlap or are subsets of more well-characterized core networks.<ref name="Uddin2019" /><br />
*Limbic<ref name="Bassett" /><ref name="Yeo" /><ref name="Bailey" /><br />
*Auditory<ref name="Yuan" /><ref name="Heine" /><br />
*Right/left executive<ref name="Yuan" /><ref name="Heine" /><br />
*Cerebellar<ref name="Bell" /><ref name="Heine" /><br />
*Spatial attention<ref name="Riedl" /><ref name="Bressler" /><br />
*Language<ref name="Bressler" /><ref name="Hutton" /><br />
*Lateral visual<ref name="Yuan" /><ref name="Bell" /><ref name="Heine" /><br />
*Temporal<ref name="Yeo" /><ref name="Shafiei" /><br />
*Visual perception/imagery<ref name="Hutton" /><br />
<br />
Different methods and data have identified several other brain networks, many of which greatly overlap or are subsets of more well-characterized core networks.<br />
*Limbic<br />
*Auditory<br />
*Right/left executive<br />
*Cerebellar<br />
*Spatial attention<br />
*Language<br />
*Lateral visual<br />
*Temporal<br />
*Visual perception/imagery<br />
<br />
其他网络不同的方法和数据已经确定了其他几个大脑网络，其中许多网络极大地重叠或者是更具特色的核心网络的子集。<br />
*边缘<br />
*听觉<br />
*右/左执行<br />
*小脑<br />
*空间注意<br />
*语言<br />
*外侧视觉<br />
*颞视知觉/图像<br />
<br />
==See also==<br />
*[[Complex network]]<br />
*[[Neural network]]<br />
<br />
*Complex network<br />
*Neural network<br />
<br />
=<nowiki>= = =</nowiki>= <br />
* 复杂网络<br />
*<br />
*神经网络<br />
*<br />
<br />
==References==<br />
{{reflist|30em}}<br />
<br />
{{Human connectomics}}<br />
<br />
[[Category:Neuroscience]]<br />
[[Category:Neural coding]]<br />
[[Category:Neural circuits]]<br />
[[Category:Neurophysiology]]<br />
<br />
Category:Neuroscience<br />
Category:Neural coding<br />
Category:Neural circuits<br />
Category:Neurophysiology<br />
<br />
类别: 神经科学类别: 神经编码类别: 神经回路类别: 神经生理学<br />
<br />
<noinclude><br />
<br />
<small>This page was moved from [[wikipedia:en:Large-scale brain network]]. Its edit history can be viewed at [[大规模脑网络/edithistory]]</small></noinclude><br />
<br />
[[Category:待整理页面]]</div>Shenky20https://wiki.swarma.org/index.php?title=%E5%A4%A7%E8%A7%84%E6%A8%A1%E8%84%91%E7%BD%91%E7%BB%9C&diff=29984大规模脑网络2022-04-04T07:50:41Z<p>Shenky20：</p>
<hr />
<div>此词条由神经动力学模型读书会词条梳理志愿者Shenky20翻译审校，翻译字数共800，未经专家审核，带来阅读不便，请见谅。<br />
<br />
{{Short description|Collections of brain regions}}<br />
'''Large-scale brain networks''' (also known as '''intrinsic brain networks''') are collections of widespread [[brain regions]] showing [[Resting state fMRI#Functional|functional connectivity]] by statistical analysis of the [[Functional magnetic resonance imaging|fMRI]] [[BOLD signal]]<ref name=Riedl>{{cite journal|last1=Riedl|first1=Valentin|last2=Utz|first2=Lukas|last3=Castrillón|first3=Gabriel|last4=Grimmer|first4=Timo|last5=Rauschecker|first5=Josef P.|last6=Ploner|first6=Markus|last7=Friston|first7=Karl J.|last8=Drzezga|first8=Alexander|last9=Sorg|first9=Christian|title=Metabolic connectivity mapping reveals effective connectivity in the resting human brain|journal=PNAS|date=January 12, 2016|volume=113|issue=2|pages=428–433|doi=10.1073/pnas.1513752113|pmid=26712010|pmc=4720331|bibcode=2016PNAS..113..428R|doi-access=free}}</ref> or other recording methods such as [[Electroencephalography|EEG]],<ref>{{Cite journal|last1=Foster|first1=Brett L.|last2=Parvizi|first2=Josef|date=2012-03-01|title=Resting oscillations and cross-frequency coupling in the human posteromedial cortex|journal=NeuroImage|volume=60|issue=1|pages=384–391|doi=10.1016/j.neuroimage.2011.12.019|pmid=22227048|issn=1053-8119|pmc=3596417}}</ref> [[Positron emission tomography|PET]]<ref>{{Cite journal|last1=Buckner|first1=Randy L.|last2=Andrews‐Hanna|first2=Jessica R.|last3=Schacter|first3=Daniel L.|date=2008|title=The Brain's Default Network|journal=Annals of the New York Academy of Sciences|language=en|volume=1124|issue=1|pages=1–38|doi=10.1196/annals.1440.011|pmid=18400922|issn=1749-6632|bibcode=2008NYASA1124....1B|s2cid=3167595}}</ref> and [[Magnetoencephalography|MEG]].<ref>{{Cite journal|last1=Morris|first1=Peter G.|last2=Smith|first2=Stephen M.|last3=Barnes|first3=Gareth R.|last4=Stephenson|first4=Mary C.|last5=Hale|first5=Joanne R.|last6=Price|first6=Darren|last7=Luckhoo|first7=Henry|last8=Woolrich|first8=Mark|last9=Brookes|first9=Matthew J.|date=2011-10-04|title=Investigating the electrophysiological basis of resting state networks using magnetoencephalography|journal=Proceedings of the National Academy of Sciences|language=en|volume=108|issue=40|pages=16783–16788|doi=10.1073/pnas.1112685108|issn=0027-8424|pmid=21930901|pmc=3189080|bibcode=2011PNAS..10816783B|doi-access=free}}</ref> An emerging paradigm in neuroscience is that cognitive tasks are performed not by individual brain regions working in isolation but by networks consisting of several discrete brain regions that are said to be "functionally connected". Functional connectivity networks may be found using algorithms such as [[cluster analysis]], spatial [[independent component analysis]] (ICA), seed based, and others.<ref name="Petersen">{{cite journal|last1=Petersen|first1=Steven|last2=Sporns|first2=Olaf|title=Brain Networks and Cognitive Architectures|journal=Neuron|date=October 2015|volume=88|issue=1|pages=207–219|doi=10.1016/j.neuron.2015.09.027|pmid=26447582|pmc=4598639 }}</ref> Synchronized brain regions may also be identified using long-range synchronization of the EEG, MEG, or other dynamic brain signals.<ref name=Bressler>{{cite journal|last1=Bressler|first1=Steven L.|last2=Menon|first2=Vinod|s2cid=5967761|title=Large scale brain networks in cognition: emerging methods and principles|journal=Trends in Cognitive Sciences|date=June 2010|volume=14|issue=6|pages=233–290|doi=10.1016/j.tics.2010.04.004|url=http://www.cell.com/trends/cognitive-sciences/issue?pii=S1364-6613(10)X0005-5|accessdate=24 January 2016|pmid=20493761}}</ref><br />
<br />
<br />
Large-scale brain networks (also known as intrinsic brain networks) are collections of widespread brain regions showing functional connectivity by statistical analysis of the fMRI BOLD signal or other recording methods such as EEG, PET and MEG. An emerging paradigm in neuroscience is that cognitive tasks are performed not by individual brain regions working in isolation but by networks consisting of several discrete brain regions that are said to be "functionally connected". Functional connectivity networks may be found using algorithms such as cluster analysis, spatial independent component analysis (ICA), seed based, and others. Synchronized brain regions may also be identified using long-range synchronization of the EEG, MEG, or other dynamic brain signals.<br />
<br />
'''<font color="#ff8000">大规模脑网络Large-scale brain networks</font>'''(也称为'''<font color="#ff8000">内在大脑网络Intrinsic brain networks</font>''')是在对基于'''<font color="#ff8000">血氧水平依赖效应BOLD</font>'''的'''<font color="#ff8000">功能性磁共振成像fMRI</font>'''信号的统计分析或其他记录方法(如'''<font color="#ff8000">脑电图EEG</font>'''、'''<font color="#ff8000">正电子发射断层扫描技术PET</font>'''和'''<font color="#ff8000">脑磁图MEG</font>''')中，表现出'''<font color="#ff8000">功能连接Functional connectivity</font>'''的'''<font color="#ff8000">脑区Brain regions</font>'''的集合。根据神经科学中一个新出现的范式，认知任务不是由单个脑区独立执行的，而是由几个互不相连的脑区“功能连接”组成的网络执行的。功能连接网络可以通过'''<font color="#ff8000">数据聚类Cluster analysis</font>'''、空间'''<font color="#ff8000">独立元素分析ICA</font>'''、种子点方法等算法来发现。同步的脑区也可以用脑电图、脑磁图或其他动态脑信号的远程同步来识别。<br />
<br />
The set of identified brain areas that are linked together in a large-scale network varies with cognitive function.<ref name="Bressler2">{{cite journal|last1=Bressler|first1=Steven L.|title=Neurocognitive networks|journal=Scholarpedia|volume=3|issue=2|pages=1567|doi=10.4249/scholarpedia.1567|year=2008|bibcode=2008SchpJ...3.1567B|doi-access=free}}</ref> When the cognitive state is not explicit (i.e., the subject is at "rest"), the large-scale brain network is a [[Resting state fMRI|resting state]] network (RSN). As a physical system with graph-like properties,<ref name="Bressler" /> a large-scale brain network has both nodes and edges and cannot be identified simply by the co-activation of brain areas. In recent decades, the analysis of brain networks was made feasible by advances in imaging techniques as well as new tools from [[graph theory]] and [[Dynamical systems theory|dynamical systems]].<br />
<br />
The set of identified brain areas that are linked together in a large-scale network varies with cognitive function. When the cognitive state is not explicit (i.e., the subject is at "rest"), the large-scale brain network is a resting state network (RSN). As a physical system with graph-like properties, a large-scale brain network has both nodes and edges and cannot be identified simply by the co-activation of brain areas. In recent decades, the analysis of brain networks was made feasible by advances in imaging techniques as well as new tools from graph theory and dynamical systems.<br />
<br />
大规模脑网络中,连接在一起的脑区集合因认知功能的不同而不同。当认知状态不明确(即主体处于“静止”状态)时，大规模脑网络是一个'''<font color="#ff8000">静息状态Resting state</font>'''网络(RSN)。作为一个具有图形特征的物理系统，大规模脑网络既有节点又有边，不能简单地通过脑区的共同激活来识别。近几十年来，成像技术不断进步，同时'''<font color="#ff8000">图论Graph theory</font>'''、'''<font color="#ff8000">动力学系统Dynamical systems</font>'''领域出现了新的技术手段，这使得脑网络分析变得可行。<br />
<br />
Large-scale brain networks are identified by their function and provide a coherent framework for understanding [[cognition]] by offering a neural model of how different cognitive functions emerge when different sets of brain regions join together as self-organized coalitions. The number and composition of the coalitions will vary with the algorithm and parameters used to identify them.<ref name="Yeo" /><ref>{{cite journal|last1=Abou Elseoud|first1=Ahmed|last2=Littow|first2=Harri|last3=Remes|first3=Jukka|last4=Starck|first4=Tuomo|last5=Nikkinen|first5=Juha|last6=Nissilä|first6=Juuso|last7=Timonen|first7=Markku|last8=Tervonen|first8=Osmo|last9=Kiviniemi1|first9=Vesa|title=Group-ICA Model Order Highlights Patterns of Functional Brain Connectivity|journal= Frontiers in Systems Neuroscience|date=2011-06-03|volume=5|pages=37|doi=10.3389/fnsys.2011.00037|pmid=21687724|pmc=3109774 |doi-access=free}}</ref> In one model, there is only the [[default mode network]] and the [[task-positive network]], but most current analyses show several networks, from a small handful to 17.<ref name="Yeo" /> The most common and stable networks are enumerated below. The regions participating in a functional network may be dynamically reconfigured.<ref name="Petersen" /><ref name="Bassett">{{cite journal|last1=Bassett|first1=Daniella|last2=Bertolero|first2=Max|title=How Matter Becomes Mind|journal=Scientific American|date=July 2019|volume=321|issue=1|page=32|url=https://www.scientificamerican.com/|accessdate=23 June 2019 }}</ref><br />
<br />
Large-scale brain networks are identified by their function and provide a coherent framework for understanding cognition by offering a neural model of how different cognitive functions emerge when different sets of brain regions join together as self-organized coalitions. The number and composition of the coalitions will vary with the algorithm and parameters used to identify them. In one model, there is only the default mode network and the task-positive network, but most current analyses show several networks, from a small handful to 17. The most common and stable networks are enumerated below. The regions participating in a functional network may be dynamically reconfigured.<br />
<br />
大规模脑网络是通过其功能来进行识别的。通过研究大规模脑网络建立神经模型，对不同脑区组合所形成的自组织联合体如何实现不同的'''<font color="#ff8000">认知Cognition</font>'''功能进行解释，就能够为认知理解提供一个连贯的框架。识别算法和参数的不同会导致所识别出的上述联合体的数量和组成有所不同。（一个模型理论认为，符合上述条件的神经模型只包含'''<font color="#ff8000">默认模式网络Default mode network</font>'''和'''<font color="#ff8000">任务激活网络Task-positive network</font>'''，但目前大多数分析理论都包括从几个到17个不等的网络。）CY下面列举了最常见且稳定的网络。'''<font color="#32CD32"> （人脑）</font>'''可以动态地重新配置参与功能网络的脑区。 <br />
<br />
Disruptions in activity in various networks have been implicated in neuropsychiatric disorders such as [[Depression (mood)|depression]], [[Alzheimer's disease|Alzheimer's]], [[Autism-spectrum disorder|autism spectrum disorder]], [[schizophrenia]], [[ADHD]]<ref>{{cite journal |last1=Griffiths |first1=Kristi R. |last2=Braund |first2=Taylor A. |last3=Kohn |first3=Michael R. |last4=Clarke |first4=Simon |last5=Williams |first5=Leanne M. |last6=Korgaonkar |first6=Mayuresh S. |title=Structural brain network topology underpinning ADHD and response to methylphenidate treatment |journal=Translational Psychiatry |date=2 March 2021 |volume=11 |issue=1 |pages=1–9 |doi=10.1038/s41398-021-01278-x | pmc=7925571 |pmid=33654073 |url=https://www.nature.com/articles/s41398-021-01278-x#citeas |access-date=16 November 2021}}</ref> and [[bipolar disorder]].<ref>{{Cite journal|url=https://www.researchgate.net/publication/51639686|title=Large-scale brain networks and psychopathology: A unifying triple network model|last=Menon|first=Vinod|s2cid=26653572|journal=Trends in Cognitive Sciences|date=2011-09-09|volume=15|issue=10|pages=483–506|doi=10.1016/j.tics.2011.08.003|pmid=21908230}}</ref><br />
<br />
Disruptions in activity in various networks have been implicated in neuropsychiatric disorders such as depression, Alzheimer's, autism spectrum disorder, schizophrenia, ADHD and bipolar disorder.<br />
<br />
各种网络活动的中断牵连到神经精神疾病，如抑郁症、老年痴呆症、自闭症光谱、精神分裂症、多动症和躁郁症。<br />
<br />
==Core networks==<br />
[[File:Heine2012x3010.png|thumb|An example that identified 10 large-scale brain networks from [[resting state fMRI]] activity through [[independent component analysis]].<ref name="Heine" />|链接=Special:FilePath/Heine2012x3010.png]]<br />
Because brain networks can be identified at various different resolutions and with various different neurobiological properties, there is no such thing as a universal atlas of brain networks that fits all circumstances.<ref>{{cite journal|last1=Eickhoff|first1=SB|last2=Yeo|first2=BTT|last3=Genon|first3=S|title=Imaging-based parcellations of the human brain.|journal=Nature Reviews. Neuroscience|date=November 2018|volume=19|issue=11|pages=672–686|doi=10.1038/s41583-018-0071-7|pmid=30305712|s2cid=52954265|url=http://juser.fz-juelich.de/record/856633/files/Eickhoff_Yeo_Genon_NRN_MainManuscriptInclFigures.pdf}}</ref> While acknowledging this problem, Uddin, Yeo, and Spreng proposed in 2019<ref name="Uddin2019">{{cite journal|last1=Uddin|first1=LQ|last2=Yeo|first2=BTT|last3=Spreng|first3=RN|title=Towards a Universal Taxonomy of Macro-scale Functional Human Brain Networks.|journal=Brain Topography|date=November 2019|volume=32|issue=6|pages=926–942|doi=10.1007/s10548-019-00744-6|pmid=31707621|pmc=7325607}}</ref> that the following six networks should be defined as core networks based on converging evidences from multiple studies<ref>{{cite journal|last1=Doucet|first1=GE|last2=Lee|first2=WH|last3=Frangou|first3=S|title=Evaluation of the spatial variability in the major resting-state networks across human brain functional atlases.|journal=Human Brain Mapping|date=2019-10-15|volume=40|issue=15|pages=4577–4587|doi=10.1002/hbm.24722|pmid=31322303|pmc=6771873}}</ref><ref name="Yeo" /><ref>{{cite journal|last1=Smith|first1=SM|last2=Fox|first2=PT|last3=Miller|first3=KL|last4=Glahn|first4=DC|last5=Fox|first5=PM|last6=Mackay|first6=CE|last7=Filippini|first7=N|last8=Watkins|first8=KE|last9=Toro|first9=R|last10=Laird|first10=AR|last11=Beckmann|first11=CF|title=Correspondence of the brain's functional architecture during activation and rest.|journal=Proceedings of the National Academy of Sciences of the United States of America|date=2009-08-04|volume=106|issue=31|pages=13040–5|doi=10.1073/pnas.0905267106|pmid=19620724|pmc=2722273|bibcode=2009PNAS..10613040S|doi-access=free}}</ref> to facilitate communication between researchers.<br />
<br />
<br />
Because brain networks can be identified at various different resolutions and with various different neurobiological properties, there is no such thing as a universal atlas of brain networks that fits all circumstances. While acknowledging this problem, Uddin, Yeo, and Spreng proposed in 2019 that the following six networks should be defined as core networks based on converging evidences from multiple studies to facilitate communication between researchers.<br />
<br />
因为大脑网络可以用不同的分辨率和不同的神经生物学特性来识别，所以没有适合所有情况的通用大脑网络图谱。在承认这个问题的同时，Uddin，Yeo，和 Spreng 在2019年提出，以下六个网络应该被定义为核心网络，基于来自多个研究的聚合证据，以促进研究人员之间的交流。<br />
<br />
===Default Mode (Medial frontoparietal)===<br />
{{Main|Default mode network}}<br />
*The default mode network is active when an individual is awake and at rest. It preferentially activates when individuals focus on internally-oriented tasks such as daydreaming, envisioning the future, retrieving memories, and [[theory of mind]]. It is negatively correlated with brain systems that focus on external visual signals. It is the most widely researched network.<ref name="Bressler" /><ref name="Bassett" /><ref>{{Cite journal|date=2012-08-15|title=The serendipitous discovery of the brain's default network|journal=NeuroImage|language=en|volume=62|issue=2|pages=1137–1145|doi=10.1016/j.neuroimage.2011.10.035|pmid=22037421|issn=1053-8119|last1=Buckner|first1=Randy L.|s2cid=9880586}}</ref><ref name="Riedl" /><ref name="Yuan"><br />
{{cite journal|last1=Yuan|first1=Rui|last2=Di|first2=Xin|last3=Taylor|first3=Paul A.|last4=Gohel|first4=Suril|last5=Tsai|first5=Yuan-Hsiung|last6=Biswal|first6=Bharat B.|title=Functional topography of the thalamocortical system in human|journal=Brain Structure and Function|date=30 April 2015|doi=10.1007/s00429-015-1018-7|pmid=25924563|pmc=6363530|volume=221|issue=4|pages=1971–1984}}</ref><ref name="Bell">{{cite journal|last1=Bell|first1=Peter T.|last2=Shine|first2=James M.|title=Estimating Large-Scale Network Convergence in the Human Functional Connectome|journal=Brain Connectivity|date=2015-11-09|volume=5|issue=9|doi=10.1089/brain.2015.0348|pmid=26005099|pages=565–74}}</ref><ref name="Heine">{{cite journal|last1=Heine|first1=Lizette|last2=Soddu|first2=Andrea|last3=Gomez|first3=Francisco|last4=Vanhaudenhuyse|first4=Audrey|last5=Tshibanda|first5=Luaba|last6=Thonnard|first6=Marie|last7=Charland-Verville|first7=Vanessa|last8=Kirsch|first8=Murielle|last9=Laureys|first9=Steven|last10=Demertzi|first10=Athena|title=Resting state networks and consciousness. Alterations of multiple resting state network connectivity in physiological, pharmacological and pathological consciousness states.|journal=Frontiers in Psychology|date=2012|volume=3|pages=295|doi=10.3389/fpsyg.2012.00295|pmid=22969735|pmc=3427917|doi-access=free}}</ref><ref name="Yeo">{{cite journal|last1=Yeo|first1=B. T. Thomas|last2=Krienen|first2=Fenna M.|last3=Sepulcre|first3=Jorge|last4=Sabuncu|first4=Mert R.|last5=Lashkari|first5=Danial|last6=Hollinshead|first6=Marisa|last7=Roffman|first7=Joshua L.|last8=Smoller|first8=Jordan W.|last9=Zöllei|first9=Lilla|last10=Polimeni|first10=Jonathan R.|last11=Fischl|first11=Bruce|last12=Liu|first12=Hesheng|last13=Buckner|first13=Randy L.|title=The organization of the human cerebral cortex estimated by intrinsic functional connectivity|journal=Journal of Neurophysiology|date=2011-09-01|volume=106|issue=3|pages=1125–1165|doi=10.1152/jn.00338.2011|pmid=21653723|pmc=3174820|bibcode=2011NatSD...2E0031H }}</ref><ref name="Shafiei">{{cite journal|last1=Shafiei|first1=Golia|last2=Zeighami|first2=Yashar|last3=Clark|first3=Crystal A.|last4=Coull|first4=Jennifer T.|last5=Nagano-Saito|first5=Atsuko|last6=Leyton|first6=Marco|last7=Dagher|first7=Alain|last8=Mišić|first8=Bratislav|title=Dopamine Signaling Modulates the Stability and Integration of Intrinsic Brain Networks|journal=Cerebral Cortex|date=2018-10-01|volume=29|issue=1|pages=397–409|doi=10.1093/cercor/bhy264|pmid=30357316|pmc=6294404 }}</ref><ref name="Bailey">{{cite journal|last1=Bailey|first1=Stephen K.|last2=Aboud|first2=Katherine S.|last3=Nguyen|first3=Tin Q.|last4=Cutting|first4=Laurie E.|title=Applying a network framework to the neurobiology of reading and dyslexia|journal=Journal of Neurodevelopmental Disorders|date=13 December 2018|volume=10|issue=1|page=37|doi=10.1186/s11689-018-9251-z|pmid=30541433|pmc=6291929 }}</ref><br />
<br />
<br />
* The default mode network is active when an individual is awake and at rest. It preferentially activates when individuals focus on internally-oriented tasks such as daydreaming, envisioning the future, retrieving memories, and theory of mind. It is negatively correlated with brain systems that focus on external visual signals. It is the most widely researched network.<br />
<br />
<br />
=<nowiki>= = 默认模式(Medial frontoparietal) = =</nowiki>= <br />
*默认模式网络在个人清醒和休息时处于活动状态。当个体专注于面向内部的任务时，比如做白日梦、展望未来、回忆和心理理论，它就会优先激活。它与专注于外部视觉信号的大脑系统负相关。它是研究最广泛的网络。<br />
<br />
===Salience (Midcingulo-Insular)===<br />
{{Main|Salience network}}<br />
*The salience network consists of several structures, including the anterior (bilateral) insula, dorsal anterior cingulate cortex, and three subcortical structures which are the ventral striatum, substantia nigra/ventral tegmental region.<ref>{{Cite journal|last1=Steimke|first1=Rosa|last2=Nomi|first2=Jason S.|last3=Calhoun|first3=Vince D.|last4=Stelzel|first4=Christine|last5=Paschke|first5=Lena M.|last6=Gaschler|first6=Robert|last7=Goschke|first7=Thomas|last8=Walter|first8=Henrik|last9=Uddin|first9=Lucina Q.|date=2017-12-01|title=Salience network dynamics underlying successful resistance of temptation|journal=Social Cognitive and Affective Neuroscience|language=en|volume=12|issue=12|pages=1928–1939|doi=10.1093/scan/nsx123|pmid=29048582|pmc=5716209|issn=1749-5016|doi-access=free}}</ref><ref name=":0">{{Citation|last=Menon|first=V.|title=Brain Mapping|chapter=Salience Network|date=2015-01-01|chapter-url=http://www.sciencedirect.com/science/article/pii/B978012397025100052X|pages=597–611|editor-last=Toga|editor-first=Arthur W.|publisher=Academic Press|isbn=978-0-12-397316-0|access-date=2019-12-08|doi=10.1016/B978-0-12-397025-1.00052-X}}</ref> It plays the key role of monitoring the [[Salience (neuroscience)|salience]] of external inputs and internal brain events.<ref name="Riedl" /><ref name="Bressler" /><ref name="Bassett" /><ref name="Yuan" /><ref name="Heine" /><ref name="Yeo" /><ref name="Shafiei" /> Specifically, it aids in directing attention by identifying important biological and cognitive events.<ref name=":0" /><ref name="Bailey" /><br />
*This network includes the ventral attention network, which primarily includes the [[temporoparietal junction]] and the ventral [[frontal cortex]] of the right hemisphere.<ref name="Uddin2019" /><ref name="Vossel" /> These areas respond when behaviorally relevant stimuli occur unexpectedly.<ref name="Vossel" /> The ventral attention network is inhibited during focused attention in which top-down processing is being used, such as when visually searching for something. This response may prevent goal-driven attention from being distracted by non-relevant stimuli. It becomes active again when the target or relevant information about the target is found.<ref name="Vossel" /><ref>{{Cite journal|last1=Shulman|first1=Gordon L.|last2=McAvoy|first2=Mark P.|last3=Cowan|first3=Melanie C.|last4=Astafiev|first4=Serguei V.|last5=Tansy|first5=Aaron P.|last6=d'Avossa|first6=Giovanni|last7=Corbetta|first7=Maurizio|date=2003-11-01|title=Quantitative Analysis of Attention and Detection Signals During Visual Search|journal=Journal of Neurophysiology|volume=90|issue=5|pages=3384–3397|doi=10.1152/jn.00343.2003|pmid=12917383|issn=0022-3077}}</ref><br />
<br />
<br />
* The salience network consists of several structures, including the anterior (bilateral) insula, dorsal anterior cingulate cortex, and three subcortical structures which are the ventral striatum, substantia nigra/ventral tegmental region. It plays the key role of monitoring the salience of external inputs and internal brain events. Specifically, it aids in directing attention by identifying important biological and cognitive events.<br />
*This network includes the ventral attention network, which primarily includes the temporoparietal junction and the ventral frontal cortex of the right hemisphere. These areas respond when behaviorally relevant stimuli occur unexpectedly. The ventral attention network is inhibited during focused attention in which top-down processing is being used, such as when visually searching for something. This response may prevent goal-driven attention from being distracted by non-relevant stimuli. It becomes active again when the target or relevant information about the target is found.<br />
<br />
突显网络由前(双)岛、前扣带回皮层和腹侧纹状体、黑质/腹侧被盖区3个皮层下结构组成。它起着监测外部输入和内部脑事件的显著性的关键作用。具体来说，它通过识别重要的生物学和认知事件来帮助引导注意力。<br />
*这个网络包括腹侧注意网络，主要包括右半球的颞顶联合区和腹侧额叶皮层。当行为相关的刺激意外发生时，这些区域会做出反应。腹侧注意网络在使用自上而下加工的集中注意过程中被抑制，例如在视觉搜索某物时。这种反应可以防止目标驱动的注意力被非相关的刺激分散。当找到目标或关于目标的相关信息时，它再次激活。<br />
<br />
===Attention (Dorsal frontoparietal)===<br />
{{Main|Dorsal attention network}}<br />
*This network is involved in the voluntary, top-down deployment of attention.<ref name="Riedl" /><ref name="Yuan" /><ref name="Bell" /><ref name="Yeo" /><ref name="Shafiei" /><ref name="Vossel">{{cite journal|last1=Vossel|first1=Simone|last2=Geng|first2=Joy J.|last3=Fink|first3=Gereon R.|title=Dorsal and Ventral Attention Systems: Distinct Neural Circuits but Collaborative Roles|journal=The Neuroscientist|date=2014|volume=20|issue=2|pages=150–159|doi=10.1177/1073858413494269|pmid=23835449|pmc=4107817}}</ref><ref name="Hutton">{{cite journal|last1=Hutton|first1=John S.|last2=Dudley|first2=Jonathan|last3=Horowitz-Kraus|first3=Tzipi|last4=DeWitt|first4=Tom|last5=Holland|first5=Scott K.|title=Functional Connectivity of Attention, Visual, and Language Networks During Audio, Illustrated, and Animated Stories in Preschool-Age Children|journal=Brain Connectivity|date=1 September 2019|volume=9|issue=7|pages=580–592|doi=10.1089/brain.2019.0679|pmid=31144523|pmc=6775495|ref=Hutton}}</ref> Within the dorsal attention network, the intraparietal sulcus and frontal eye fields influence the visual areas of the brain. These influencing factors allow for the orientation of attention.<ref>{{Cite journal|last1=Fox|first1=Michael D.|last2=Corbetta|first2=Maurizio|last3=Snyder|first3=Abraham Z.|last4=Vincent|first4=Justin L.|last5=Raichle|first5=Marcus E.|date=2006-06-27|title=Spontaneous neuronal activity distinguishes human dorsal and ventral attention systems|journal=Proceedings of the National Academy of Sciences|language=en|volume=103|issue=26|pages=10046–10051|doi=10.1073/pnas.0604187103|issn=0027-8424|pmid=16788060|pmc=1480402|bibcode=2006PNAS..10310046F|doi-access=free}}</ref><ref name="Vossel" /><ref name="Bailey" /><br />
<br />
<br />
*This network is involved in the voluntary, top-down deployment of attention. Within the dorsal attention network, the intraparietal sulcus and frontal eye fields influence the visual areas of the brain. These influencing factors allow for the orientation of attention.<br />
<br />
=<nowiki>= = 注意力(背侧额顶骨) = =</nowiki>= <br />
*这个网络参与了自发的、自上而下的注意力分配。在背侧注意网络中，顶内沟和额眼影响大脑的视觉区域。这些影响因素决定了注意力的方向。<br />
<br />
===Control (Lateral frontoparietal)===<br />
{{Main|Frontoparietal network}}<br />
*This network initiates and modulates cognitive control and comprises 18 sub-regions of the brain.<ref>{{Cite journal|last1=Scolari|first1=Miranda|last2=Seidl-Rathkopf|first2=Katharina N|last3=Kastner|first3=Sabine|date=2015-02-01|title=Functions of the human frontoparietal attention network: Evidence from neuroimaging|journal=Current Opinion in Behavioral Sciences|series=Cognitive control|volume=1|pages=32–39|doi=10.1016/j.cobeha.2014.08.003|issn=2352-1546|pmid=27398396|pmc=4936532}}</ref> There is a strong correlation between fluid intelligence and the involvement of the fronto-parietal network with other networks.<ref>{{Cite journal|last1=Marek|first1=Scott|last2=Dosenbach|first2=Nico U. F.|date=June 2018|title=The frontoparietal network: function, electrophysiology, and importance of individual precision mapping|journal=Dialogues in Clinical Neuroscience|volume=20|issue=2|pages=133–140|doi=10.31887/DCNS.2018.20.2/smarek|issn=1294-8322|pmc=6136121|pmid=30250390}}</ref><br />
<br />
<br />
*This network initiates and modulates cognitive control and comprises 18 sub-regions of the brain. There is a strong correlation between fluid intelligence and the involvement of the fronto-parietal network with other networks.<br />
<br />
=<nowiki>= = 控制(侧额顶骨) = =</nowiki>= <br />
*这个网络启动和调节认知控制，包括大脑的18个亚区。在流体智力和额顶网络与其他网络的参与之间有很强的相关性。<br />
<br />
*Versions of this network have also been called the central executive (or executive control) network and the cognitive control network.<ref name="Uddin2019" /><br />
<br />
*Versions of this network have also been called the central executive (or executive control) network and the cognitive control network.<br />
<br />
<br />
* 这种网络的版本也被称为中央执行(或执行控制)网络和认知控制网络。<br />
<br />
===Sensorimotor or Somatomotor (Pericentral)===<br />
{{Main|Sensorimotor network}}<br />
*This network processes somatosensory information and coordinates motion.<ref name="Heine" /><ref name="Yeo" /><ref name="Shafiei" /><ref name="Bassett" /><ref name="Yuan" /> The [[auditory cortex]] may be included.<ref name="Uddin2019" /><ref name="Yeo" /><br />
<br />
<br />
*This network processes somatosensory information and coordinates motion. The auditory cortex may be included.<br />
<br />
===Sensorimotor or Somatomotor (Pericentral)===<br />
<br />
<br />
*This network processes somatosensory information and coordinates motion.听觉皮层可能也包括在内。<br />
<br />
===Visual (Occipital)===<br />
{{See|Visual cortex}}<br />
* This network handles visual information processing.<ref name="Yang">{{cite journal|last1=Yang|first1=Yan-li|last2=Deng|first2=Hong-xia|last3=Xing|first3=Gui-yang|last4=Xia|first4=Xiao-luan|last5=Li|first5=Hai-fang|title=Brain functional network connectivity based on a visual task: visual information processing-related brain regions are significantly activated in the task state|journal=Neural Regeneration Research|date=2015|volume=10|issue=2|pages=298–307|doi=10.4103/1673-5374.152386|pmid=25883631|pmc=4392680 }}</ref><br />
<br />
<br />
*This network handles visual information processing.<br />
<br />
= = = = = = = = = = = = = 这个网络处理视觉信息。<br />
<br />
==Other networks==<br />
Different methods and data have identified several other brain networks, many of which greatly overlap or are subsets of more well-characterized core networks.<ref name="Uddin2019" /><br />
*Limbic<ref name="Bassett" /><ref name="Yeo" /><ref name="Bailey" /><br />
*Auditory<ref name="Yuan" /><ref name="Heine" /><br />
*Right/left executive<ref name="Yuan" /><ref name="Heine" /><br />
*Cerebellar<ref name="Bell" /><ref name="Heine" /><br />
*Spatial attention<ref name="Riedl" /><ref name="Bressler" /><br />
*Language<ref name="Bressler" /><ref name="Hutton" /><br />
*Lateral visual<ref name="Yuan" /><ref name="Bell" /><ref name="Heine" /><br />
*Temporal<ref name="Yeo" /><ref name="Shafiei" /><br />
*Visual perception/imagery<ref name="Hutton" /><br />
<br />
Different methods and data have identified several other brain networks, many of which greatly overlap or are subsets of more well-characterized core networks.<br />
*Limbic<br />
*Auditory<br />
*Right/left executive<br />
*Cerebellar<br />
*Spatial attention<br />
* Language<br />
*Lateral visual<br />
*Temporal<br />
*Visual perception/imagery<br />
<br />
其他网络不同的方法和数据已经确定了其他几个大脑网络，其中许多网络极大地重叠或者是更具特色的核心网络的子集。<br />
* 边缘<br />
*听觉<br />
*右/左执行<br />
*小脑<br />
*空间注意<br />
*语言<br />
*外侧视觉<br />
*颞视知觉/图像<br />
<br />
==See also==<br />
*[[Complex network]]<br />
*[[Neural network]]<br />
<br />
*Complex network<br />
*Neural network<br />
<br />
=<nowiki>= = =</nowiki>= <br />
*复杂网络<br />
*<br />
*神经网络<br />
*<br />
<br />
==References==<br />
{{reflist|30em}}<br />
<br />
{{Human connectomics}}<br />
<br />
[[Category:Neuroscience]]<br />
[[Category:Neural coding]]<br />
[[Category:Neural circuits]]<br />
[[Category:Neurophysiology]]<br />
<br />
Category:Neuroscience<br />
Category:Neural coding<br />
Category:Neural circuits<br />
Category:Neurophysiology<br />
<br />
类别: 神经科学类别: 神经编码类别: 神经回路类别: 神经生理学<br />
<br />
<noinclude><br />
<br />
<small>This page was moved from [[wikipedia:en:Large-scale brain network]]. Its edit history can be viewed at [[大规模脑网络/edithistory]]</small></noinclude><br />
<br />
[[Category:待整理页面]]</div>Shenky20https://wiki.swarma.org/index.php?title=%E5%A4%A7%E8%A7%84%E6%A8%A1%E8%84%91%E7%BD%91%E7%BB%9C&diff=29983大规模脑网络2022-04-04T07:46:43Z<p>Shenky20：</p>
<hr />
<div>此词条由神经动力学模型读书会词条梳理志愿者Shenky20翻译审校，翻译字数共800，未经专家审核，带来阅读不便，请见谅。<br />
<br />
{{Short description|Collections of brain regions}}<br />
'''Large-scale brain networks''' (also known as '''intrinsic brain networks''') are collections of widespread [[brain regions]] showing [[Resting state fMRI#Functional|functional connectivity]] by statistical analysis of the [[Functional magnetic resonance imaging|fMRI]] [[BOLD signal]]<ref name=Riedl>{{cite journal|last1=Riedl|first1=Valentin|last2=Utz|first2=Lukas|last3=Castrillón|first3=Gabriel|last4=Grimmer|first4=Timo|last5=Rauschecker|first5=Josef P.|last6=Ploner|first6=Markus|last7=Friston|first7=Karl J.|last8=Drzezga|first8=Alexander|last9=Sorg|first9=Christian|title=Metabolic connectivity mapping reveals effective connectivity in the resting human brain|journal=PNAS|date=January 12, 2016|volume=113|issue=2|pages=428–433|doi=10.1073/pnas.1513752113|pmid=26712010|pmc=4720331|bibcode=2016PNAS..113..428R|doi-access=free}}</ref> or other recording methods such as [[Electroencephalography|EEG]],<ref>{{Cite journal|last1=Foster|first1=Brett L.|last2=Parvizi|first2=Josef|date=2012-03-01|title=Resting oscillations and cross-frequency coupling in the human posteromedial cortex|journal=NeuroImage|volume=60|issue=1|pages=384–391|doi=10.1016/j.neuroimage.2011.12.019|pmid=22227048|issn=1053-8119|pmc=3596417}}</ref> [[Positron emission tomography|PET]]<ref>{{Cite journal|last1=Buckner|first1=Randy L.|last2=Andrews‐Hanna|first2=Jessica R.|last3=Schacter|first3=Daniel L.|date=2008|title=The Brain's Default Network|journal=Annals of the New York Academy of Sciences|language=en|volume=1124|issue=1|pages=1–38|doi=10.1196/annals.1440.011|pmid=18400922|issn=1749-6632|bibcode=2008NYASA1124....1B|s2cid=3167595}}</ref> and [[Magnetoencephalography|MEG]].<ref>{{Cite journal|last1=Morris|first1=Peter G.|last2=Smith|first2=Stephen M.|last3=Barnes|first3=Gareth R.|last4=Stephenson|first4=Mary C.|last5=Hale|first5=Joanne R.|last6=Price|first6=Darren|last7=Luckhoo|first7=Henry|last8=Woolrich|first8=Mark|last9=Brookes|first9=Matthew J.|date=2011-10-04|title=Investigating the electrophysiological basis of resting state networks using magnetoencephalography|journal=Proceedings of the National Academy of Sciences|language=en|volume=108|issue=40|pages=16783–16788|doi=10.1073/pnas.1112685108|issn=0027-8424|pmid=21930901|pmc=3189080|bibcode=2011PNAS..10816783B|doi-access=free}}</ref> An emerging paradigm in neuroscience is that cognitive tasks are performed not by individual brain regions working in isolation but by networks consisting of several discrete brain regions that are said to be "functionally connected". Functional connectivity networks may be found using algorithms such as [[cluster analysis]], spatial [[independent component analysis]] (ICA), seed based, and others.<ref name="Petersen">{{cite journal|last1=Petersen|first1=Steven|last2=Sporns|first2=Olaf|title=Brain Networks and Cognitive Architectures|journal=Neuron|date=October 2015|volume=88|issue=1|pages=207–219|doi=10.1016/j.neuron.2015.09.027|pmid=26447582|pmc=4598639 }}</ref> Synchronized brain regions may also be identified using long-range synchronization of the EEG, MEG, or other dynamic brain signals.<ref name=Bressler>{{cite journal|last1=Bressler|first1=Steven L.|last2=Menon|first2=Vinod|s2cid=5967761|title=Large scale brain networks in cognition: emerging methods and principles|journal=Trends in Cognitive Sciences|date=June 2010|volume=14|issue=6|pages=233–290|doi=10.1016/j.tics.2010.04.004|url=http://www.cell.com/trends/cognitive-sciences/issue?pii=S1364-6613(10)X0005-5|accessdate=24 January 2016|pmid=20493761}}</ref><br />
<br />
<br />
Large-scale brain networks (also known as intrinsic brain networks) are collections of widespread brain regions showing functional connectivity by statistical analysis of the fMRI BOLD signal or other recording methods such as EEG, PET and MEG. An emerging paradigm in neuroscience is that cognitive tasks are performed not by individual brain regions working in isolation but by networks consisting of several discrete brain regions that are said to be "functionally connected". Functional connectivity networks may be found using algorithms such as cluster analysis, spatial independent component analysis (ICA), seed based, and others. Synchronized brain regions may also be identified using long-range synchronization of the EEG, MEG, or other dynamic brain signals.<br />
<br />
'''<font color="#ff8000">大规模脑网络Large-scale brain networks</font>'''(也称为'''<font color="#ff8000">内在大脑网络Intrinsic brain networks</font>''')是在对基于'''<font color="#ff8000">血氧水平依赖效应BOLD</font>'''的'''<font color="#ff8000">功能性磁共振成像fMRI</font>'''信号的统计分析或其他记录方法(如'''<font color="#ff8000">脑电图EEG</font>'''、'''<font color="#ff8000">正电子发射断层扫描技术PET</font>'''和'''<font color="#ff8000">脑磁图MEG</font>''')中，表现出'''<font color="#ff8000">功能连接Functional connectivity</font>'''的'''<font color="#ff8000">脑区Brain regions</font>'''的集合。根据神经科学中一个新出现的范式，认知任务不是由单个脑区独立执行的，而是由几个互不相连的脑区“功能连接”组成的网络执行的。功能连接网络可以通过'''<font color="#ff8000">数据聚类Cluster analysis</font>'''、空间'''<font color="#ff8000">独立元素分析ICA</font>'''、种子点方法等算法来发现。同步的脑区也可以用脑电图、脑磁图或其他动态脑信号的远程同步来识别。<br />
<br />
The set of identified brain areas that are linked together in a large-scale network varies with cognitive function.<ref name="Bressler2">{{cite journal|last1=Bressler|first1=Steven L.|title=Neurocognitive networks|journal=Scholarpedia|volume=3|issue=2|pages=1567|doi=10.4249/scholarpedia.1567|year=2008|bibcode=2008SchpJ...3.1567B|doi-access=free}}</ref> When the cognitive state is not explicit (i.e., the subject is at "rest"), the large-scale brain network is a [[Resting state fMRI|resting state]] network (RSN). As a physical system with graph-like properties,<ref name="Bressler" /> a large-scale brain network has both nodes and edges and cannot be identified simply by the co-activation of brain areas. In recent decades, the analysis of brain networks was made feasible by advances in imaging techniques as well as new tools from [[graph theory]] and [[Dynamical systems theory|dynamical systems]].<br />
<br />
The set of identified brain areas that are linked together in a large-scale network varies with cognitive function. When the cognitive state is not explicit (i.e., the subject is at "rest"), the large-scale brain network is a resting state network (RSN). As a physical system with graph-like properties, a large-scale brain network has both nodes and edges and cannot be identified simply by the co-activation of brain areas. In recent decades, the analysis of brain networks was made feasible by advances in imaging techniques as well as new tools from graph theory and dynamical systems.<br />
<br />
大规模脑网络中,连接在一起的脑区集合因认知功能的不同而不同。当认知状态不明确(即主体处于“静止”状态)时，大规模脑网络是一个'''<font color="#ff8000">静息状态Resting state</font>'''网络(RSN)。作为一个具有图形特征的物理系统，大规模脑网络既有节点又有边，不能简单地通过脑区的共同激活来识别。近几十年来，成像技术不断进步，同时'''<font color="#ff8000">图论Graph theory</font>'''、'''<font color="#ff8000">动力学系统Dynamical systems</font>'''领域出现了新的技术手段，这使得脑网络分析变得可行。<br />
<br />
Large-scale brain networks are identified by their function and provide a coherent framework for understanding [[cognition]] by offering a neural model of how different cognitive functions emerge when different sets of brain regions join together as self-organized coalitions. The number and composition of the coalitions will vary with the algorithm and parameters used to identify them.<ref name="Yeo" /><ref>{{cite journal|last1=Abou Elseoud|first1=Ahmed|last2=Littow|first2=Harri|last3=Remes|first3=Jukka|last4=Starck|first4=Tuomo|last5=Nikkinen|first5=Juha|last6=Nissilä|first6=Juuso|last7=Timonen|first7=Markku|last8=Tervonen|first8=Osmo|last9=Kiviniemi1|first9=Vesa|title=Group-ICA Model Order Highlights Patterns of Functional Brain Connectivity|journal= Frontiers in Systems Neuroscience|date=2011-06-03|volume=5|pages=37|doi=10.3389/fnsys.2011.00037|pmid=21687724|pmc=3109774 |doi-access=free}}</ref> In one model, there is only the [[default mode network]] and the [[task-positive network]], but most current analyses show several networks, from a small handful to 17.<ref name="Yeo" /> The most common and stable networks are enumerated below. The regions participating in a functional network may be dynamically reconfigured.<ref name="Petersen" /><ref name="Bassett">{{cite journal|last1=Bassett|first1=Daniella|last2=Bertolero|first2=Max|title=How Matter Becomes Mind|journal=Scientific American|date=July 2019|volume=321|issue=1|page=32|url=https://www.scientificamerican.com/|accessdate=23 June 2019 }}</ref><br />
<br />
Large-scale brain networks are identified by their function and provide a coherent framework for understanding cognition by offering a neural model of how different cognitive functions emerge when different sets of brain regions join together as self-organized coalitions. The number and composition of the coalitions will vary with the algorithm and parameters used to identify them. In one model, there is only the default mode network and the task-positive network, but most current analyses show several networks, from a small handful to 17. The most common and stable networks are enumerated below. The regions participating in a functional network may be dynamically reconfigured.<br />
<br />
大规模的大脑网络是通过其功能识别的，并通过提供一个神经模型，说明当不同的大脑区域组合在一起形成自组织的联盟时，不同的认知功能是如何产生的，从而为理解认知提供了一个连贯的框架。联盟的数量和组成将根据识别联盟的算法和参数而有所不同。在一个模型中，只有默认模式网络和任务正向网络，但目前大多数分析显示，从少数到17个网络。下面列举了最常见和最稳定的网络。可以动态地重新配置参与功能网络的区域。<br />
<br />
Disruptions in activity in various networks have been implicated in neuropsychiatric disorders such as [[Depression (mood)|depression]], [[Alzheimer's disease|Alzheimer's]], [[Autism-spectrum disorder|autism spectrum disorder]], [[schizophrenia]], [[ADHD]]<ref>{{cite journal |last1=Griffiths |first1=Kristi R. |last2=Braund |first2=Taylor A. |last3=Kohn |first3=Michael R. |last4=Clarke |first4=Simon |last5=Williams |first5=Leanne M. |last6=Korgaonkar |first6=Mayuresh S. |title=Structural brain network topology underpinning ADHD and response to methylphenidate treatment |journal=Translational Psychiatry |date=2 March 2021 |volume=11 |issue=1 |pages=1–9 |doi=10.1038/s41398-021-01278-x | pmc=7925571 |pmid=33654073 |url=https://www.nature.com/articles/s41398-021-01278-x#citeas |access-date=16 November 2021}}</ref> and [[bipolar disorder]].<ref>{{Cite journal|url=https://www.researchgate.net/publication/51639686|title=Large-scale brain networks and psychopathology: A unifying triple network model|last=Menon|first=Vinod|s2cid=26653572|journal=Trends in Cognitive Sciences|date=2011-09-09|volume=15|issue=10|pages=483–506|doi=10.1016/j.tics.2011.08.003|pmid=21908230}}</ref><br />
<br />
Disruptions in activity in various networks have been implicated in neuropsychiatric disorders such as depression, Alzheimer's, autism spectrum disorder, schizophrenia, ADHD and bipolar disorder.<br />
<br />
各种网络活动的中断牵连到神经精神疾病，如抑郁症、老年痴呆症、自闭症光谱、精神分裂症、多动症和躁郁症。<br />
<br />
==Core networks==<br />
[[File:Heine2012x3010.png|thumb|An example that identified 10 large-scale brain networks from [[resting state fMRI]] activity through [[independent component analysis]].<ref name="Heine" />|链接=Special:FilePath/Heine2012x3010.png]]<br />
Because brain networks can be identified at various different resolutions and with various different neurobiological properties, there is no such thing as a universal atlas of brain networks that fits all circumstances.<ref>{{cite journal|last1=Eickhoff|first1=SB|last2=Yeo|first2=BTT|last3=Genon|first3=S|title=Imaging-based parcellations of the human brain.|journal=Nature Reviews. Neuroscience|date=November 2018|volume=19|issue=11|pages=672–686|doi=10.1038/s41583-018-0071-7|pmid=30305712|s2cid=52954265|url=http://juser.fz-juelich.de/record/856633/files/Eickhoff_Yeo_Genon_NRN_MainManuscriptInclFigures.pdf}}</ref> While acknowledging this problem, Uddin, Yeo, and Spreng proposed in 2019<ref name="Uddin2019">{{cite journal|last1=Uddin|first1=LQ|last2=Yeo|first2=BTT|last3=Spreng|first3=RN|title=Towards a Universal Taxonomy of Macro-scale Functional Human Brain Networks.|journal=Brain Topography|date=November 2019|volume=32|issue=6|pages=926–942|doi=10.1007/s10548-019-00744-6|pmid=31707621|pmc=7325607}}</ref> that the following six networks should be defined as core networks based on converging evidences from multiple studies<ref>{{cite journal|last1=Doucet|first1=GE|last2=Lee|first2=WH|last3=Frangou|first3=S|title=Evaluation of the spatial variability in the major resting-state networks across human brain functional atlases.|journal=Human Brain Mapping|date=2019-10-15|volume=40|issue=15|pages=4577–4587|doi=10.1002/hbm.24722|pmid=31322303|pmc=6771873}}</ref><ref name="Yeo" /><ref>{{cite journal|last1=Smith|first1=SM|last2=Fox|first2=PT|last3=Miller|first3=KL|last4=Glahn|first4=DC|last5=Fox|first5=PM|last6=Mackay|first6=CE|last7=Filippini|first7=N|last8=Watkins|first8=KE|last9=Toro|first9=R|last10=Laird|first10=AR|last11=Beckmann|first11=CF|title=Correspondence of the brain's functional architecture during activation and rest.|journal=Proceedings of the National Academy of Sciences of the United States of America|date=2009-08-04|volume=106|issue=31|pages=13040–5|doi=10.1073/pnas.0905267106|pmid=19620724|pmc=2722273|bibcode=2009PNAS..10613040S|doi-access=free}}</ref> to facilitate communication between researchers.<br />
<br />
<br />
Because brain networks can be identified at various different resolutions and with various different neurobiological properties, there is no such thing as a universal atlas of brain networks that fits all circumstances. While acknowledging this problem, Uddin, Yeo, and Spreng proposed in 2019 that the following six networks should be defined as core networks based on converging evidences from multiple studies to facilitate communication between researchers.<br />
<br />
因为大脑网络可以用不同的分辨率和不同的神经生物学特性来识别，所以没有适合所有情况的通用大脑网络图谱。在承认这个问题的同时，Uddin，Yeo，和 Spreng 在2019年提出，以下六个网络应该被定义为核心网络，基于来自多个研究的聚合证据，以促进研究人员之间的交流。<br />
<br />
===Default Mode (Medial frontoparietal)===<br />
{{Main|Default mode network}}<br />
*The default mode network is active when an individual is awake and at rest. It preferentially activates when individuals focus on internally-oriented tasks such as daydreaming, envisioning the future, retrieving memories, and [[theory of mind]]. It is negatively correlated with brain systems that focus on external visual signals. It is the most widely researched network.<ref name="Bressler" /><ref name="Bassett" /><ref>{{Cite journal|date=2012-08-15|title=The serendipitous discovery of the brain's default network|journal=NeuroImage|language=en|volume=62|issue=2|pages=1137–1145|doi=10.1016/j.neuroimage.2011.10.035|pmid=22037421|issn=1053-8119|last1=Buckner|first1=Randy L.|s2cid=9880586}}</ref><ref name="Riedl" /><ref name="Yuan"><br />
{{cite journal|last1=Yuan|first1=Rui|last2=Di|first2=Xin|last3=Taylor|first3=Paul A.|last4=Gohel|first4=Suril|last5=Tsai|first5=Yuan-Hsiung|last6=Biswal|first6=Bharat B.|title=Functional topography of the thalamocortical system in human|journal=Brain Structure and Function|date=30 April 2015|doi=10.1007/s00429-015-1018-7|pmid=25924563|pmc=6363530|volume=221|issue=4|pages=1971–1984}}</ref><ref name="Bell">{{cite journal|last1=Bell|first1=Peter T.|last2=Shine|first2=James M.|title=Estimating Large-Scale Network Convergence in the Human Functional Connectome|journal=Brain Connectivity|date=2015-11-09|volume=5|issue=9|doi=10.1089/brain.2015.0348|pmid=26005099|pages=565–74}}</ref><ref name="Heine">{{cite journal|last1=Heine|first1=Lizette|last2=Soddu|first2=Andrea|last3=Gomez|first3=Francisco|last4=Vanhaudenhuyse|first4=Audrey|last5=Tshibanda|first5=Luaba|last6=Thonnard|first6=Marie|last7=Charland-Verville|first7=Vanessa|last8=Kirsch|first8=Murielle|last9=Laureys|first9=Steven|last10=Demertzi|first10=Athena|title=Resting state networks and consciousness. Alterations of multiple resting state network connectivity in physiological, pharmacological and pathological consciousness states.|journal=Frontiers in Psychology|date=2012|volume=3|pages=295|doi=10.3389/fpsyg.2012.00295|pmid=22969735|pmc=3427917|doi-access=free}}</ref><ref name="Yeo">{{cite journal|last1=Yeo|first1=B. T. Thomas|last2=Krienen|first2=Fenna M.|last3=Sepulcre|first3=Jorge|last4=Sabuncu|first4=Mert R.|last5=Lashkari|first5=Danial|last6=Hollinshead|first6=Marisa|last7=Roffman|first7=Joshua L.|last8=Smoller|first8=Jordan W.|last9=Zöllei|first9=Lilla|last10=Polimeni|first10=Jonathan R.|last11=Fischl|first11=Bruce|last12=Liu|first12=Hesheng|last13=Buckner|first13=Randy L.|title=The organization of the human cerebral cortex estimated by intrinsic functional connectivity|journal=Journal of Neurophysiology|date=2011-09-01|volume=106|issue=3|pages=1125–1165|doi=10.1152/jn.00338.2011|pmid=21653723|pmc=3174820|bibcode=2011NatSD...2E0031H }}</ref><ref name="Shafiei">{{cite journal|last1=Shafiei|first1=Golia|last2=Zeighami|first2=Yashar|last3=Clark|first3=Crystal A.|last4=Coull|first4=Jennifer T.|last5=Nagano-Saito|first5=Atsuko|last6=Leyton|first6=Marco|last7=Dagher|first7=Alain|last8=Mišić|first8=Bratislav|title=Dopamine Signaling Modulates the Stability and Integration of Intrinsic Brain Networks|journal=Cerebral Cortex|date=2018-10-01|volume=29|issue=1|pages=397–409|doi=10.1093/cercor/bhy264|pmid=30357316|pmc=6294404 }}</ref><ref name="Bailey">{{cite journal|last1=Bailey|first1=Stephen K.|last2=Aboud|first2=Katherine S.|last3=Nguyen|first3=Tin Q.|last4=Cutting|first4=Laurie E.|title=Applying a network framework to the neurobiology of reading and dyslexia|journal=Journal of Neurodevelopmental Disorders|date=13 December 2018|volume=10|issue=1|page=37|doi=10.1186/s11689-018-9251-z|pmid=30541433|pmc=6291929 }}</ref><br />
<br />
<br />
*The default mode network is active when an individual is awake and at rest. It preferentially activates when individuals focus on internally-oriented tasks such as daydreaming, envisioning the future, retrieving memories, and theory of mind. It is negatively correlated with brain systems that focus on external visual signals. It is the most widely researched network.<br />
<br />
<br />
=<nowiki>= = 默认模式(Medial frontoparietal) = =</nowiki>= <br />
*默认模式网络在个人清醒和休息时处于活动状态。当个体专注于面向内部的任务时，比如做白日梦、展望未来、回忆和心理理论，它就会优先激活。它与专注于外部视觉信号的大脑系统负相关。它是研究最广泛的网络。<br />
<br />
=== Salience (Midcingulo-Insular)===<br />
{{Main|Salience network}}<br />
*The salience network consists of several structures, including the anterior (bilateral) insula, dorsal anterior cingulate cortex, and three subcortical structures which are the ventral striatum, substantia nigra/ventral tegmental region.<ref>{{Cite journal|last1=Steimke|first1=Rosa|last2=Nomi|first2=Jason S.|last3=Calhoun|first3=Vince D.|last4=Stelzel|first4=Christine|last5=Paschke|first5=Lena M.|last6=Gaschler|first6=Robert|last7=Goschke|first7=Thomas|last8=Walter|first8=Henrik|last9=Uddin|first9=Lucina Q.|date=2017-12-01|title=Salience network dynamics underlying successful resistance of temptation|journal=Social Cognitive and Affective Neuroscience|language=en|volume=12|issue=12|pages=1928–1939|doi=10.1093/scan/nsx123|pmid=29048582|pmc=5716209|issn=1749-5016|doi-access=free}}</ref><ref name=":0">{{Citation|last=Menon|first=V.|title=Brain Mapping|chapter=Salience Network|date=2015-01-01|chapter-url=http://www.sciencedirect.com/science/article/pii/B978012397025100052X|pages=597–611|editor-last=Toga|editor-first=Arthur W.|publisher=Academic Press|isbn=978-0-12-397316-0|access-date=2019-12-08|doi=10.1016/B978-0-12-397025-1.00052-X}}</ref> It plays the key role of monitoring the [[Salience (neuroscience)|salience]] of external inputs and internal brain events.<ref name="Riedl" /><ref name="Bressler" /><ref name="Bassett" /><ref name="Yuan" /><ref name="Heine" /><ref name="Yeo" /><ref name="Shafiei" /> Specifically, it aids in directing attention by identifying important biological and cognitive events.<ref name=":0" /><ref name="Bailey" /><br />
* This network includes the ventral attention network, which primarily includes the [[temporoparietal junction]] and the ventral [[frontal cortex]] of the right hemisphere.<ref name="Uddin2019" /><ref name="Vossel" /> These areas respond when behaviorally relevant stimuli occur unexpectedly.<ref name="Vossel" /> The ventral attention network is inhibited during focused attention in which top-down processing is being used, such as when visually searching for something. This response may prevent goal-driven attention from being distracted by non-relevant stimuli. It becomes active again when the target or relevant information about the target is found.<ref name="Vossel" /><ref>{{Cite journal|last1=Shulman|first1=Gordon L.|last2=McAvoy|first2=Mark P.|last3=Cowan|first3=Melanie C.|last4=Astafiev|first4=Serguei V.|last5=Tansy|first5=Aaron P.|last6=d'Avossa|first6=Giovanni|last7=Corbetta|first7=Maurizio|date=2003-11-01|title=Quantitative Analysis of Attention and Detection Signals During Visual Search|journal=Journal of Neurophysiology|volume=90|issue=5|pages=3384–3397|doi=10.1152/jn.00343.2003|pmid=12917383|issn=0022-3077}}</ref><br />
<br />
<br />
*The salience network consists of several structures, including the anterior (bilateral) insula, dorsal anterior cingulate cortex, and three subcortical structures which are the ventral striatum, substantia nigra/ventral tegmental region. It plays the key role of monitoring the salience of external inputs and internal brain events. Specifically, it aids in directing attention by identifying important biological and cognitive events.<br />
*This network includes the ventral attention network, which primarily includes the temporoparietal junction and the ventral frontal cortex of the right hemisphere. These areas respond when behaviorally relevant stimuli occur unexpectedly. The ventral attention network is inhibited during focused attention in which top-down processing is being used, such as when visually searching for something. This response may prevent goal-driven attention from being distracted by non-relevant stimuli. It becomes active again when the target or relevant information about the target is found.<br />
<br />
突显网络由前(双)岛、前扣带回皮层和腹侧纹状体、黑质/腹侧被盖区3个皮层下结构组成。它起着监测外部输入和内部脑事件的显著性的关键作用。具体来说，它通过识别重要的生物学和认知事件来帮助引导注意力。<br />
*这个网络包括腹侧注意网络，主要包括右半球的颞顶联合区和腹侧额叶皮层。当行为相关的刺激意外发生时，这些区域会做出反应。腹侧注意网络在使用自上而下加工的集中注意过程中被抑制，例如在视觉搜索某物时。这种反应可以防止目标驱动的注意力被非相关的刺激分散。当找到目标或关于目标的相关信息时，它再次激活。<br />
<br />
===Attention (Dorsal frontoparietal)===<br />
{{Main|Dorsal attention network}}<br />
*This network is involved in the voluntary, top-down deployment of attention.<ref name="Riedl" /><ref name="Yuan" /><ref name="Bell" /><ref name="Yeo" /><ref name="Shafiei" /><ref name="Vossel">{{cite journal|last1=Vossel|first1=Simone|last2=Geng|first2=Joy J.|last3=Fink|first3=Gereon R.|title=Dorsal and Ventral Attention Systems: Distinct Neural Circuits but Collaborative Roles|journal=The Neuroscientist|date=2014|volume=20|issue=2|pages=150–159|doi=10.1177/1073858413494269|pmid=23835449|pmc=4107817}}</ref><ref name="Hutton">{{cite journal|last1=Hutton|first1=John S.|last2=Dudley|first2=Jonathan|last3=Horowitz-Kraus|first3=Tzipi|last4=DeWitt|first4=Tom|last5=Holland|first5=Scott K.|title=Functional Connectivity of Attention, Visual, and Language Networks During Audio, Illustrated, and Animated Stories in Preschool-Age Children|journal=Brain Connectivity|date=1 September 2019|volume=9|issue=7|pages=580–592|doi=10.1089/brain.2019.0679|pmid=31144523|pmc=6775495|ref=Hutton}}</ref> Within the dorsal attention network, the intraparietal sulcus and frontal eye fields influence the visual areas of the brain. These influencing factors allow for the orientation of attention.<ref>{{Cite journal|last1=Fox|first1=Michael D.|last2=Corbetta|first2=Maurizio|last3=Snyder|first3=Abraham Z.|last4=Vincent|first4=Justin L.|last5=Raichle|first5=Marcus E.|date=2006-06-27|title=Spontaneous neuronal activity distinguishes human dorsal and ventral attention systems|journal=Proceedings of the National Academy of Sciences|language=en|volume=103|issue=26|pages=10046–10051|doi=10.1073/pnas.0604187103|issn=0027-8424|pmid=16788060|pmc=1480402|bibcode=2006PNAS..10310046F|doi-access=free}}</ref><ref name="Vossel" /><ref name="Bailey" /><br />
<br />
<br />
*This network is involved in the voluntary, top-down deployment of attention. Within the dorsal attention network, the intraparietal sulcus and frontal eye fields influence the visual areas of the brain. These influencing factors allow for the orientation of attention.<br />
<br />
=<nowiki>= = 注意力(背侧额顶骨) = =</nowiki>= <br />
*这个网络参与了自发的、自上而下的注意力分配。在背侧注意网络中，顶内沟和额眼影响大脑的视觉区域。这些影响因素决定了注意力的方向。<br />
<br />
===Control (Lateral frontoparietal)===<br />
{{Main|Frontoparietal network}}<br />
*This network initiates and modulates cognitive control and comprises 18 sub-regions of the brain.<ref>{{Cite journal|last1=Scolari|first1=Miranda|last2=Seidl-Rathkopf|first2=Katharina N|last3=Kastner|first3=Sabine|date=2015-02-01|title=Functions of the human frontoparietal attention network: Evidence from neuroimaging|journal=Current Opinion in Behavioral Sciences|series=Cognitive control|volume=1|pages=32–39|doi=10.1016/j.cobeha.2014.08.003|issn=2352-1546|pmid=27398396|pmc=4936532}}</ref> There is a strong correlation between fluid intelligence and the involvement of the fronto-parietal network with other networks.<ref>{{Cite journal|last1=Marek|first1=Scott|last2=Dosenbach|first2=Nico U. F.|date=June 2018|title=The frontoparietal network: function, electrophysiology, and importance of individual precision mapping|journal=Dialogues in Clinical Neuroscience|volume=20|issue=2|pages=133–140|doi=10.31887/DCNS.2018.20.2/smarek|issn=1294-8322|pmc=6136121|pmid=30250390}}</ref><br />
<br />
<br />
*This network initiates and modulates cognitive control and comprises 18 sub-regions of the brain. There is a strong correlation between fluid intelligence and the involvement of the fronto-parietal network with other networks.<br />
<br />
=<nowiki>= = 控制(侧额顶骨) = =</nowiki>= <br />
*这个网络启动和调节认知控制，包括大脑的18个亚区。在流体智力和额顶网络与其他网络的参与之间有很强的相关性。<br />
<br />
*Versions of this network have also been called the central executive (or executive control) network and the cognitive control network.<ref name="Uddin2019" /><br />
<br />
*Versions of this network have also been called the central executive (or executive control) network and the cognitive control network.<br />
<br />
<br />
*这种网络的版本也被称为中央执行(或执行控制)网络和认知控制网络。<br />
<br />
===Sensorimotor or Somatomotor (Pericentral)===<br />
{{Main|Sensorimotor network}}<br />
*This network processes somatosensory information and coordinates motion.<ref name="Heine" /><ref name="Yeo" /><ref name="Shafiei" /><ref name="Bassett" /><ref name="Yuan" /> The [[auditory cortex]] may be included.<ref name="Uddin2019" /><ref name="Yeo" /><br />
<br />
<br />
*This network processes somatosensory information and coordinates motion. The auditory cortex may be included.<br />
<br />
===Sensorimotor or Somatomotor (Pericentral)===<br />
<br />
<br />
*This network processes somatosensory information and coordinates motion.听觉皮层可能也包括在内。<br />
<br />
===Visual (Occipital)===<br />
{{See|Visual cortex}}<br />
*This network handles visual information processing.<ref name="Yang">{{cite journal|last1=Yang|first1=Yan-li|last2=Deng|first2=Hong-xia|last3=Xing|first3=Gui-yang|last4=Xia|first4=Xiao-luan|last5=Li|first5=Hai-fang|title=Brain functional network connectivity based on a visual task: visual information processing-related brain regions are significantly activated in the task state|journal=Neural Regeneration Research|date=2015|volume=10|issue=2|pages=298–307|doi=10.4103/1673-5374.152386|pmid=25883631|pmc=4392680 }}</ref><br />
<br />
<br />
*This network handles visual information processing.<br />
<br />
= = = = = = = = = = = = = 这个网络处理视觉信息。<br />
<br />
==Other networks ==<br />
Different methods and data have identified several other brain networks, many of which greatly overlap or are subsets of more well-characterized core networks.<ref name="Uddin2019" /><br />
*Limbic<ref name="Bassett" /><ref name="Yeo" /><ref name="Bailey" /><br />
*Auditory<ref name="Yuan" /><ref name="Heine" /><br />
*Right/left executive<ref name="Yuan" /><ref name="Heine" /><br />
* Cerebellar<ref name="Bell" /><ref name="Heine" /><br />
*Spatial attention<ref name="Riedl" /><ref name="Bressler" /><br />
*Language<ref name="Bressler" /><ref name="Hutton" /><br />
*Lateral visual<ref name="Yuan" /><ref name="Bell" /><ref name="Heine" /><br />
* Temporal<ref name="Yeo" /><ref name="Shafiei" /><br />
*Visual perception/imagery<ref name="Hutton" /><br />
<br />
Different methods and data have identified several other brain networks, many of which greatly overlap or are subsets of more well-characterized core networks.<br />
*Limbic<br />
*Auditory<br />
*Right/left executive<br />
*Cerebellar<br />
*Spatial attention<br />
*Language<br />
*Lateral visual<br />
*Temporal<br />
*Visual perception/imagery<br />
<br />
其他网络不同的方法和数据已经确定了其他几个大脑网络，其中许多网络极大地重叠或者是更具特色的核心网络的子集。<br />
*边缘<br />
*听觉<br />
*右/左执行<br />
*小脑<br />
*空间注意<br />
*语言<br />
* 外侧视觉<br />
*颞视知觉/图像<br />
<br />
== See also==<br />
*[[Complex network]]<br />
*[[Neural network]]<br />
<br />
*Complex network<br />
*Neural network<br />
<br />
=<nowiki>= = =</nowiki>= <br />
*复杂网络<br />
*<br />
*神经网络<br />
*<br />
<br />
==References==<br />
{{reflist|30em}}<br />
<br />
{{Human connectomics}}<br />
<br />
[[Category:Neuroscience]]<br />
[[Category:Neural coding]]<br />
[[Category:Neural circuits]]<br />
[[Category:Neurophysiology]]<br />
<br />
Category:Neuroscience<br />
Category:Neural coding<br />
Category:Neural circuits<br />
Category:Neurophysiology<br />
<br />
类别: 神经科学类别: 神经编码类别: 神经回路类别: 神经生理学<br />
<br />
<noinclude><br />
<br />
<small>This page was moved from [[wikipedia:en:Large-scale brain network]]. Its edit history can be viewed at [[大规模脑网络/edithistory]]</small></noinclude><br />
<br />
[[Category:待整理页面]]</div>Shenky20https://wiki.swarma.org/index.php?title=%E5%A4%A7%E8%A7%84%E6%A8%A1%E8%84%91%E7%BD%91%E7%BB%9C&diff=29982大规模脑网络2022-04-04T07:43:36Z<p>Shenky20：</p>
<hr />
<div>此词条由神经动力学模型读书会词条梳理志愿者Shenky20翻译审校，翻译字数共800，未经专家审核，带来阅读不便，请见谅。<br />
<br />
{{Short description|Collections of brain regions}}<br />
'''Large-scale brain networks''' (also known as '''intrinsic brain networks''') are collections of widespread [[brain regions]] showing [[Resting state fMRI#Functional|functional connectivity]] by statistical analysis of the [[Functional magnetic resonance imaging|fMRI]] [[BOLD signal]]<ref name=Riedl>{{cite journal|last1=Riedl|first1=Valentin|last2=Utz|first2=Lukas|last3=Castrillón|first3=Gabriel|last4=Grimmer|first4=Timo|last5=Rauschecker|first5=Josef P.|last6=Ploner|first6=Markus|last7=Friston|first7=Karl J.|last8=Drzezga|first8=Alexander|last9=Sorg|first9=Christian|title=Metabolic connectivity mapping reveals effective connectivity in the resting human brain|journal=PNAS|date=January 12, 2016|volume=113|issue=2|pages=428–433|doi=10.1073/pnas.1513752113|pmid=26712010|pmc=4720331|bibcode=2016PNAS..113..428R|doi-access=free}}</ref> or other recording methods such as [[Electroencephalography|EEG]],<ref>{{Cite journal|last1=Foster|first1=Brett L.|last2=Parvizi|first2=Josef|date=2012-03-01|title=Resting oscillations and cross-frequency coupling in the human posteromedial cortex|journal=NeuroImage|volume=60|issue=1|pages=384–391|doi=10.1016/j.neuroimage.2011.12.019|pmid=22227048|issn=1053-8119|pmc=3596417}}</ref> [[Positron emission tomography|PET]]<ref>{{Cite journal|last1=Buckner|first1=Randy L.|last2=Andrews‐Hanna|first2=Jessica R.|last3=Schacter|first3=Daniel L.|date=2008|title=The Brain's Default Network|journal=Annals of the New York Academy of Sciences|language=en|volume=1124|issue=1|pages=1–38|doi=10.1196/annals.1440.011|pmid=18400922|issn=1749-6632|bibcode=2008NYASA1124....1B|s2cid=3167595}}</ref> and [[Magnetoencephalography|MEG]].<ref>{{Cite journal|last1=Morris|first1=Peter G.|last2=Smith|first2=Stephen M.|last3=Barnes|first3=Gareth R.|last4=Stephenson|first4=Mary C.|last5=Hale|first5=Joanne R.|last6=Price|first6=Darren|last7=Luckhoo|first7=Henry|last8=Woolrich|first8=Mark|last9=Brookes|first9=Matthew J.|date=2011-10-04|title=Investigating the electrophysiological basis of resting state networks using magnetoencephalography|journal=Proceedings of the National Academy of Sciences|language=en|volume=108|issue=40|pages=16783–16788|doi=10.1073/pnas.1112685108|issn=0027-8424|pmid=21930901|pmc=3189080|bibcode=2011PNAS..10816783B|doi-access=free}}</ref> An emerging paradigm in neuroscience is that cognitive tasks are performed not by individual brain regions working in isolation but by networks consisting of several discrete brain regions that are said to be "functionally connected". Functional connectivity networks may be found using algorithms such as [[cluster analysis]], spatial [[independent component analysis]] (ICA), seed based, and others.<ref name="Petersen">{{cite journal|last1=Petersen|first1=Steven|last2=Sporns|first2=Olaf|title=Brain Networks and Cognitive Architectures|journal=Neuron|date=October 2015|volume=88|issue=1|pages=207–219|doi=10.1016/j.neuron.2015.09.027|pmid=26447582|pmc=4598639 }}</ref> Synchronized brain regions may also be identified using long-range synchronization of the EEG, MEG, or other dynamic brain signals.<ref name=Bressler>{{cite journal|last1=Bressler|first1=Steven L.|last2=Menon|first2=Vinod|s2cid=5967761|title=Large scale brain networks in cognition: emerging methods and principles|journal=Trends in Cognitive Sciences|date=June 2010|volume=14|issue=6|pages=233–290|doi=10.1016/j.tics.2010.04.004|url=http://www.cell.com/trends/cognitive-sciences/issue?pii=S1364-6613(10)X0005-5|accessdate=24 January 2016|pmid=20493761}}</ref><br />
<br />
<br />
Large-scale brain networks (also known as intrinsic brain networks) are collections of widespread brain regions showing functional connectivity by statistical analysis of the fMRI BOLD signal or other recording methods such as EEG, PET and MEG. An emerging paradigm in neuroscience is that cognitive tasks are performed not by individual brain regions working in isolation but by networks consisting of several discrete brain regions that are said to be "functionally connected". Functional connectivity networks may be found using algorithms such as cluster analysis, spatial independent component analysis (ICA), seed based, and others. Synchronized brain regions may also be identified using long-range synchronization of the EEG, MEG, or other dynamic brain signals.<br />
<br />
'''<font color="#ff8000">大规模脑网络Large-scale brain networks</font>'''(也称为'''<font color="#ff8000">内在大脑网络Intrinsic brain networks</font>''')是在对基于'''<font color="#ff8000">血氧水平依赖效应BOLD</font>'''的'''<font color="#ff8000">功能性磁共振成像fMRI</font>'''信号的统计分析或其他记录方法(如'''<font color="#ff8000">脑电图EEG</font>'''、'''<font color="#ff8000">正电子发射断层扫描技术PET</font>'''和'''<font color="#ff8000">脑磁图MEG</font>''')中，表现出'''<font color="#ff8000">功能连接Functional connectivity</font>'''的'''<font color="#ff8000">脑区Brain regions</font>'''的集合。根据神经科学中一个新出现的范式，认知任务不是由单个脑区独立执行的，而是由几个互不相连的脑区“功能连接”组成的网络执行的。功能连接网络可以通过'''<font color="#ff8000">数据聚类Cluster analysis</font>'''、空间'''<font color="#ff8000">独立元素分析ICA</font>'''、种子点方法等算法来发现。同步的脑区也可以用脑电图、脑磁图或其他动态脑信号的远程同步来识别。<br />
<br />
The set of identified brain areas that are linked together in a large-scale network varies with cognitive function.<ref name="Bressler2">{{cite journal|last1=Bressler|first1=Steven L.|title=Neurocognitive networks|journal=Scholarpedia|volume=3|issue=2|pages=1567|doi=10.4249/scholarpedia.1567|year=2008|bibcode=2008SchpJ...3.1567B|doi-access=free}}</ref> When the cognitive state is not explicit (i.e., the subject is at "rest"), the large-scale brain network is a [[Resting state fMRI|resting state]] network (RSN). As a physical system with graph-like properties,<ref name="Bressler" /> a large-scale brain network has both nodes and edges and cannot be identified simply by the co-activation of brain areas. In recent decades, the analysis of brain networks was made feasible by advances in imaging techniques as well as new tools from [[graph theory]] and [[Dynamical systems theory|dynamical systems]].<br />
<br />
The set of identified brain areas that are linked together in a large-scale network varies with cognitive function. When the cognitive state is not explicit (i.e., the subject is at "rest"), the large-scale brain network is a resting state network (RSN). As a physical system with graph-like properties, a large-scale brain network has both nodes and edges and cannot be identified simply by the co-activation of brain areas. In recent decades, the analysis of brain networks was made feasible by advances in imaging techniques as well as new tools from graph theory and dynamical systems.<br />
<br />
在一个大规模的网络中，连接在一起的大脑区域的集合因认知功能的不同而不同。当认知状态不明确(即主体处于“静止”状态)时，大规模的脑网络是一个静息状态网络(RSN)。作为一个具有图形特征的物理系统，大规模的脑网络既有节点又有边，不能简单地通过脑区的共同激活来识别。近几十年来，由于成像技术的进步以及图论和动力学系统的新工具，脑网络的分析变得可行。<br />
<br />
Large-scale brain networks are identified by their function and provide a coherent framework for understanding [[cognition]] by offering a neural model of how different cognitive functions emerge when different sets of brain regions join together as self-organized coalitions. The number and composition of the coalitions will vary with the algorithm and parameters used to identify them.<ref name="Yeo" /><ref>{{cite journal|last1=Abou Elseoud|first1=Ahmed|last2=Littow|first2=Harri|last3=Remes|first3=Jukka|last4=Starck|first4=Tuomo|last5=Nikkinen|first5=Juha|last6=Nissilä|first6=Juuso|last7=Timonen|first7=Markku|last8=Tervonen|first8=Osmo|last9=Kiviniemi1|first9=Vesa|title=Group-ICA Model Order Highlights Patterns of Functional Brain Connectivity|journal= Frontiers in Systems Neuroscience|date=2011-06-03|volume=5|pages=37|doi=10.3389/fnsys.2011.00037|pmid=21687724|pmc=3109774 |doi-access=free}}</ref> In one model, there is only the [[default mode network]] and the [[task-positive network]], but most current analyses show several networks, from a small handful to 17.<ref name="Yeo" /> The most common and stable networks are enumerated below. The regions participating in a functional network may be dynamically reconfigured.<ref name="Petersen" /><ref name="Bassett">{{cite journal|last1=Bassett|first1=Daniella|last2=Bertolero|first2=Max|title=How Matter Becomes Mind|journal=Scientific American|date=July 2019|volume=321|issue=1|page=32|url=https://www.scientificamerican.com/|accessdate=23 June 2019 }}</ref><br />
<br />
Large-scale brain networks are identified by their function and provide a coherent framework for understanding cognition by offering a neural model of how different cognitive functions emerge when different sets of brain regions join together as self-organized coalitions. The number and composition of the coalitions will vary with the algorithm and parameters used to identify them. In one model, there is only the default mode network and the task-positive network, but most current analyses show several networks, from a small handful to 17. The most common and stable networks are enumerated below. The regions participating in a functional network may be dynamically reconfigured.<br />
<br />
大规模的大脑网络是通过其功能识别的，并通过提供一个神经模型，说明当不同的大脑区域组合在一起形成自组织的联盟时，不同的认知功能是如何产生的，从而为理解认知提供了一个连贯的框架。联盟的数量和组成将根据识别联盟的算法和参数而有所不同。在一个模型中，只有默认模式网络和任务正向网络，但目前大多数分析显示，从少数到17个网络。下面列举了最常见和最稳定的网络。可以动态地重新配置参与功能网络的区域。<br />
<br />
Disruptions in activity in various networks have been implicated in neuropsychiatric disorders such as [[Depression (mood)|depression]], [[Alzheimer's disease|Alzheimer's]], [[Autism-spectrum disorder|autism spectrum disorder]], [[schizophrenia]], [[ADHD]]<ref>{{cite journal |last1=Griffiths |first1=Kristi R. |last2=Braund |first2=Taylor A. |last3=Kohn |first3=Michael R. |last4=Clarke |first4=Simon |last5=Williams |first5=Leanne M. |last6=Korgaonkar |first6=Mayuresh S. |title=Structural brain network topology underpinning ADHD and response to methylphenidate treatment |journal=Translational Psychiatry |date=2 March 2021 |volume=11 |issue=1 |pages=1–9 |doi=10.1038/s41398-021-01278-x | pmc=7925571 |pmid=33654073 |url=https://www.nature.com/articles/s41398-021-01278-x#citeas |access-date=16 November 2021}}</ref> and [[bipolar disorder]].<ref>{{Cite journal|url=https://www.researchgate.net/publication/51639686|title=Large-scale brain networks and psychopathology: A unifying triple network model|last=Menon|first=Vinod|s2cid=26653572|journal=Trends in Cognitive Sciences|date=2011-09-09|volume=15|issue=10|pages=483–506|doi=10.1016/j.tics.2011.08.003|pmid=21908230}}</ref><br />
<br />
Disruptions in activity in various networks have been implicated in neuropsychiatric disorders such as depression, Alzheimer's, autism spectrum disorder, schizophrenia, ADHD and bipolar disorder.<br />
<br />
各种网络活动的中断牵连到神经精神疾病，如抑郁症、老年痴呆症、自闭症光谱、精神分裂症、多动症和躁郁症。<br />
<br />
==Core networks ==<br />
[[File:Heine2012x3010.png|thumb|An example that identified 10 large-scale brain networks from [[resting state fMRI]] activity through [[independent component analysis]].<ref name="Heine" />|链接=Special:FilePath/Heine2012x3010.png]]<br />
Because brain networks can be identified at various different resolutions and with various different neurobiological properties, there is no such thing as a universal atlas of brain networks that fits all circumstances.<ref>{{cite journal|last1=Eickhoff|first1=SB|last2=Yeo|first2=BTT|last3=Genon|first3=S|title=Imaging-based parcellations of the human brain.|journal=Nature Reviews. Neuroscience|date=November 2018|volume=19|issue=11|pages=672–686|doi=10.1038/s41583-018-0071-7|pmid=30305712|s2cid=52954265|url=http://juser.fz-juelich.de/record/856633/files/Eickhoff_Yeo_Genon_NRN_MainManuscriptInclFigures.pdf}}</ref> While acknowledging this problem, Uddin, Yeo, and Spreng proposed in 2019<ref name="Uddin2019">{{cite journal|last1=Uddin|first1=LQ|last2=Yeo|first2=BTT|last3=Spreng|first3=RN|title=Towards a Universal Taxonomy of Macro-scale Functional Human Brain Networks.|journal=Brain Topography|date=November 2019|volume=32|issue=6|pages=926–942|doi=10.1007/s10548-019-00744-6|pmid=31707621|pmc=7325607}}</ref> that the following six networks should be defined as core networks based on converging evidences from multiple studies<ref>{{cite journal|last1=Doucet|first1=GE|last2=Lee|first2=WH|last3=Frangou|first3=S|title=Evaluation of the spatial variability in the major resting-state networks across human brain functional atlases.|journal=Human Brain Mapping|date=2019-10-15|volume=40|issue=15|pages=4577–4587|doi=10.1002/hbm.24722|pmid=31322303|pmc=6771873}}</ref><ref name="Yeo" /><ref>{{cite journal|last1=Smith|first1=SM|last2=Fox|first2=PT|last3=Miller|first3=KL|last4=Glahn|first4=DC|last5=Fox|first5=PM|last6=Mackay|first6=CE|last7=Filippini|first7=N|last8=Watkins|first8=KE|last9=Toro|first9=R|last10=Laird|first10=AR|last11=Beckmann|first11=CF|title=Correspondence of the brain's functional architecture during activation and rest.|journal=Proceedings of the National Academy of Sciences of the United States of America|date=2009-08-04|volume=106|issue=31|pages=13040–5|doi=10.1073/pnas.0905267106|pmid=19620724|pmc=2722273|bibcode=2009PNAS..10613040S|doi-access=free}}</ref> to facilitate communication between researchers.<br />
<br />
<br />
Because brain networks can be identified at various different resolutions and with various different neurobiological properties, there is no such thing as a universal atlas of brain networks that fits all circumstances. While acknowledging this problem, Uddin, Yeo, and Spreng proposed in 2019 that the following six networks should be defined as core networks based on converging evidences from multiple studies to facilitate communication between researchers.<br />
<br />
因为大脑网络可以用不同的分辨率和不同的神经生物学特性来识别，所以没有适合所有情况的通用大脑网络图谱。在承认这个问题的同时，Uddin，Yeo，和 Spreng 在2019年提出，以下六个网络应该被定义为核心网络，基于来自多个研究的聚合证据，以促进研究人员之间的交流。<br />
<br />
===Default Mode (Medial frontoparietal)===<br />
{{Main|Default mode network}}<br />
*The default mode network is active when an individual is awake and at rest. It preferentially activates when individuals focus on internally-oriented tasks such as daydreaming, envisioning the future, retrieving memories, and [[theory of mind]]. It is negatively correlated with brain systems that focus on external visual signals. It is the most widely researched network.<ref name="Bressler" /><ref name="Bassett" /><ref>{{Cite journal|date=2012-08-15|title=The serendipitous discovery of the brain's default network|journal=NeuroImage|language=en|volume=62|issue=2|pages=1137–1145|doi=10.1016/j.neuroimage.2011.10.035|pmid=22037421|issn=1053-8119|last1=Buckner|first1=Randy L.|s2cid=9880586}}</ref><ref name="Riedl" /><ref name="Yuan"><br />
{{cite journal|last1=Yuan|first1=Rui|last2=Di|first2=Xin|last3=Taylor|first3=Paul A.|last4=Gohel|first4=Suril|last5=Tsai|first5=Yuan-Hsiung|last6=Biswal|first6=Bharat B.|title=Functional topography of the thalamocortical system in human|journal=Brain Structure and Function|date=30 April 2015|doi=10.1007/s00429-015-1018-7|pmid=25924563|pmc=6363530|volume=221|issue=4|pages=1971–1984}}</ref><ref name="Bell">{{cite journal|last1=Bell|first1=Peter T.|last2=Shine|first2=James M.|title=Estimating Large-Scale Network Convergence in the Human Functional Connectome|journal=Brain Connectivity|date=2015-11-09|volume=5|issue=9|doi=10.1089/brain.2015.0348|pmid=26005099|pages=565–74}}</ref><ref name="Heine">{{cite journal|last1=Heine|first1=Lizette|last2=Soddu|first2=Andrea|last3=Gomez|first3=Francisco|last4=Vanhaudenhuyse|first4=Audrey|last5=Tshibanda|first5=Luaba|last6=Thonnard|first6=Marie|last7=Charland-Verville|first7=Vanessa|last8=Kirsch|first8=Murielle|last9=Laureys|first9=Steven|last10=Demertzi|first10=Athena|title=Resting state networks and consciousness. Alterations of multiple resting state network connectivity in physiological, pharmacological and pathological consciousness states.|journal=Frontiers in Psychology|date=2012|volume=3|pages=295|doi=10.3389/fpsyg.2012.00295|pmid=22969735|pmc=3427917|doi-access=free}}</ref><ref name="Yeo">{{cite journal|last1=Yeo|first1=B. T. Thomas|last2=Krienen|first2=Fenna M.|last3=Sepulcre|first3=Jorge|last4=Sabuncu|first4=Mert R.|last5=Lashkari|first5=Danial|last6=Hollinshead|first6=Marisa|last7=Roffman|first7=Joshua L.|last8=Smoller|first8=Jordan W.|last9=Zöllei|first9=Lilla|last10=Polimeni|first10=Jonathan R.|last11=Fischl|first11=Bruce|last12=Liu|first12=Hesheng|last13=Buckner|first13=Randy L.|title=The organization of the human cerebral cortex estimated by intrinsic functional connectivity|journal=Journal of Neurophysiology|date=2011-09-01|volume=106|issue=3|pages=1125–1165|doi=10.1152/jn.00338.2011|pmid=21653723|pmc=3174820|bibcode=2011NatSD...2E0031H }}</ref><ref name="Shafiei">{{cite journal|last1=Shafiei|first1=Golia|last2=Zeighami|first2=Yashar|last3=Clark|first3=Crystal A.|last4=Coull|first4=Jennifer T.|last5=Nagano-Saito|first5=Atsuko|last6=Leyton|first6=Marco|last7=Dagher|first7=Alain|last8=Mišić|first8=Bratislav|title=Dopamine Signaling Modulates the Stability and Integration of Intrinsic Brain Networks|journal=Cerebral Cortex|date=2018-10-01|volume=29|issue=1|pages=397–409|doi=10.1093/cercor/bhy264|pmid=30357316|pmc=6294404 }}</ref><ref name="Bailey">{{cite journal|last1=Bailey|first1=Stephen K.|last2=Aboud|first2=Katherine S.|last3=Nguyen|first3=Tin Q.|last4=Cutting|first4=Laurie E.|title=Applying a network framework to the neurobiology of reading and dyslexia|journal=Journal of Neurodevelopmental Disorders|date=13 December 2018|volume=10|issue=1|page=37|doi=10.1186/s11689-018-9251-z|pmid=30541433|pmc=6291929 }}</ref><br />
<br />
<br />
*The default mode network is active when an individual is awake and at rest. It preferentially activates when individuals focus on internally-oriented tasks such as daydreaming, envisioning the future, retrieving memories, and theory of mind. It is negatively correlated with brain systems that focus on external visual signals. It is the most widely researched network.<br />
<br />
<br />
=<nowiki>= = 默认模式(Medial frontoparietal) = =</nowiki>= <br />
*默认模式网络在个人清醒和休息时处于活动状态。当个体专注于面向内部的任务时，比如做白日梦、展望未来、回忆和心理理论，它就会优先激活。它与专注于外部视觉信号的大脑系统负相关。它是研究最广泛的网络。<br />
<br />
=== Salience (Midcingulo-Insular)===<br />
{{Main|Salience network}}<br />
*The salience network consists of several structures, including the anterior (bilateral) insula, dorsal anterior cingulate cortex, and three subcortical structures which are the ventral striatum, substantia nigra/ventral tegmental region.<ref>{{Cite journal|last1=Steimke|first1=Rosa|last2=Nomi|first2=Jason S.|last3=Calhoun|first3=Vince D.|last4=Stelzel|first4=Christine|last5=Paschke|first5=Lena M.|last6=Gaschler|first6=Robert|last7=Goschke|first7=Thomas|last8=Walter|first8=Henrik|last9=Uddin|first9=Lucina Q.|date=2017-12-01|title=Salience network dynamics underlying successful resistance of temptation|journal=Social Cognitive and Affective Neuroscience|language=en|volume=12|issue=12|pages=1928–1939|doi=10.1093/scan/nsx123|pmid=29048582|pmc=5716209|issn=1749-5016|doi-access=free}}</ref><ref name=":0">{{Citation|last=Menon|first=V.|title=Brain Mapping|chapter=Salience Network|date=2015-01-01|chapter-url=http://www.sciencedirect.com/science/article/pii/B978012397025100052X|pages=597–611|editor-last=Toga|editor-first=Arthur W.|publisher=Academic Press|isbn=978-0-12-397316-0|access-date=2019-12-08|doi=10.1016/B978-0-12-397025-1.00052-X}}</ref> It plays the key role of monitoring the [[Salience (neuroscience)|salience]] of external inputs and internal brain events.<ref name="Riedl" /><ref name="Bressler" /><ref name="Bassett" /><ref name="Yuan" /><ref name="Heine" /><ref name="Yeo" /><ref name="Shafiei" /> Specifically, it aids in directing attention by identifying important biological and cognitive events.<ref name=":0" /><ref name="Bailey" /><br />
*This network includes the ventral attention network, which primarily includes the [[temporoparietal junction]] and the ventral [[frontal cortex]] of the right hemisphere.<ref name="Uddin2019" /><ref name="Vossel" /> These areas respond when behaviorally relevant stimuli occur unexpectedly.<ref name="Vossel" /> The ventral attention network is inhibited during focused attention in which top-down processing is being used, such as when visually searching for something. This response may prevent goal-driven attention from being distracted by non-relevant stimuli. It becomes active again when the target or relevant information about the target is found.<ref name="Vossel" /><ref>{{Cite journal|last1=Shulman|first1=Gordon L.|last2=McAvoy|first2=Mark P.|last3=Cowan|first3=Melanie C.|last4=Astafiev|first4=Serguei V.|last5=Tansy|first5=Aaron P.|last6=d'Avossa|first6=Giovanni|last7=Corbetta|first7=Maurizio|date=2003-11-01|title=Quantitative Analysis of Attention and Detection Signals During Visual Search|journal=Journal of Neurophysiology|volume=90|issue=5|pages=3384–3397|doi=10.1152/jn.00343.2003|pmid=12917383|issn=0022-3077}}</ref><br />
<br />
<br />
*The salience network consists of several structures, including the anterior (bilateral) insula, dorsal anterior cingulate cortex, and three subcortical structures which are the ventral striatum, substantia nigra/ventral tegmental region. It plays the key role of monitoring the salience of external inputs and internal brain events. Specifically, it aids in directing attention by identifying important biological and cognitive events.<br />
*This network includes the ventral attention network, which primarily includes the temporoparietal junction and the ventral frontal cortex of the right hemisphere. These areas respond when behaviorally relevant stimuli occur unexpectedly. The ventral attention network is inhibited during focused attention in which top-down processing is being used, such as when visually searching for something. This response may prevent goal-driven attention from being distracted by non-relevant stimuli. It becomes active again when the target or relevant information about the target is found.<br />
<br />
突显网络由前(双)岛、前扣带回皮层和腹侧纹状体、黑质/腹侧被盖区3个皮层下结构组成。它起着监测外部输入和内部脑事件的显著性的关键作用。具体来说，它通过识别重要的生物学和认知事件来帮助引导注意力。<br />
*这个网络包括腹侧注意网络，主要包括右半球的颞顶联合区和腹侧额叶皮层。当行为相关的刺激意外发生时，这些区域会做出反应。腹侧注意网络在使用自上而下加工的集中注意过程中被抑制，例如在视觉搜索某物时。这种反应可以防止目标驱动的注意力被非相关的刺激分散。当找到目标或关于目标的相关信息时，它再次激活。<br />
<br />
===Attention (Dorsal frontoparietal)===<br />
{{Main|Dorsal attention network}}<br />
*This network is involved in the voluntary, top-down deployment of attention.<ref name="Riedl" /><ref name="Yuan" /><ref name="Bell" /><ref name="Yeo" /><ref name="Shafiei" /><ref name="Vossel">{{cite journal|last1=Vossel|first1=Simone|last2=Geng|first2=Joy J.|last3=Fink|first3=Gereon R.|title=Dorsal and Ventral Attention Systems: Distinct Neural Circuits but Collaborative Roles|journal=The Neuroscientist|date=2014|volume=20|issue=2|pages=150–159|doi=10.1177/1073858413494269|pmid=23835449|pmc=4107817}}</ref><ref name="Hutton">{{cite journal|last1=Hutton|first1=John S.|last2=Dudley|first2=Jonathan|last3=Horowitz-Kraus|first3=Tzipi|last4=DeWitt|first4=Tom|last5=Holland|first5=Scott K.|title=Functional Connectivity of Attention, Visual, and Language Networks During Audio, Illustrated, and Animated Stories in Preschool-Age Children|journal=Brain Connectivity|date=1 September 2019|volume=9|issue=7|pages=580–592|doi=10.1089/brain.2019.0679|pmid=31144523|pmc=6775495|ref=Hutton}}</ref> Within the dorsal attention network, the intraparietal sulcus and frontal eye fields influence the visual areas of the brain. These influencing factors allow for the orientation of attention.<ref>{{Cite journal|last1=Fox|first1=Michael D.|last2=Corbetta|first2=Maurizio|last3=Snyder|first3=Abraham Z.|last4=Vincent|first4=Justin L.|last5=Raichle|first5=Marcus E.|date=2006-06-27|title=Spontaneous neuronal activity distinguishes human dorsal and ventral attention systems|journal=Proceedings of the National Academy of Sciences|language=en|volume=103|issue=26|pages=10046–10051|doi=10.1073/pnas.0604187103|issn=0027-8424|pmid=16788060|pmc=1480402|bibcode=2006PNAS..10310046F|doi-access=free}}</ref><ref name="Vossel" /><ref name="Bailey" /><br />
<br />
<br />
*This network is involved in the voluntary, top-down deployment of attention. Within the dorsal attention network, the intraparietal sulcus and frontal eye fields influence the visual areas of the brain. These influencing factors allow for the orientation of attention.<br />
<br />
=<nowiki>= = 注意力(背侧额顶骨) = =</nowiki>= <br />
*这个网络参与了自发的、自上而下的注意力分配。在背侧注意网络中，顶内沟和额眼影响大脑的视觉区域。这些影响因素决定了注意力的方向。<br />
<br />
===Control (Lateral frontoparietal) ===<br />
{{Main|Frontoparietal network}}<br />
*This network initiates and modulates cognitive control and comprises 18 sub-regions of the brain.<ref>{{Cite journal|last1=Scolari|first1=Miranda|last2=Seidl-Rathkopf|first2=Katharina N|last3=Kastner|first3=Sabine|date=2015-02-01|title=Functions of the human frontoparietal attention network: Evidence from neuroimaging|journal=Current Opinion in Behavioral Sciences|series=Cognitive control|volume=1|pages=32–39|doi=10.1016/j.cobeha.2014.08.003|issn=2352-1546|pmid=27398396|pmc=4936532}}</ref> There is a strong correlation between fluid intelligence and the involvement of the fronto-parietal network with other networks.<ref>{{Cite journal|last1=Marek|first1=Scott|last2=Dosenbach|first2=Nico U. F.|date=June 2018|title=The frontoparietal network: function, electrophysiology, and importance of individual precision mapping|journal=Dialogues in Clinical Neuroscience|volume=20|issue=2|pages=133–140|doi=10.31887/DCNS.2018.20.2/smarek|issn=1294-8322|pmc=6136121|pmid=30250390}}</ref><br />
<br />
<br />
*This network initiates and modulates cognitive control and comprises 18 sub-regions of the brain. There is a strong correlation between fluid intelligence and the involvement of the fronto-parietal network with other networks.<br />
<br />
=<nowiki>= = 控制(侧额顶骨) = =</nowiki>= <br />
*这个网络启动和调节认知控制，包括大脑的18个亚区。在流体智力和额顶网络与其他网络的参与之间有很强的相关性。<br />
<br />
*Versions of this network have also been called the central executive (or executive control) network and the cognitive control network.<ref name="Uddin2019" /><br />
<br />
*Versions of this network have also been called the central executive (or executive control) network and the cognitive control network.<br />
<br />
<br />
*这种网络的版本也被称为中央执行(或执行控制)网络和认知控制网络。<br />
<br />
===Sensorimotor or Somatomotor (Pericentral)===<br />
{{Main|Sensorimotor network}}<br />
*This network processes somatosensory information and coordinates motion.<ref name="Heine" /><ref name="Yeo" /><ref name="Shafiei" /><ref name="Bassett" /><ref name="Yuan" /> The [[auditory cortex]] may be included.<ref name="Uddin2019" /><ref name="Yeo" /><br />
<br />
<br />
*This network processes somatosensory information and coordinates motion. The auditory cortex may be included.<br />
<br />
===Sensorimotor or Somatomotor (Pericentral)===<br />
<br />
<br />
*This network processes somatosensory information and coordinates motion.听觉皮层可能也包括在内。<br />
<br />
===Visual (Occipital)===<br />
{{See|Visual cortex}}<br />
*This network handles visual information processing.<ref name="Yang">{{cite journal|last1=Yang|first1=Yan-li|last2=Deng|first2=Hong-xia|last3=Xing|first3=Gui-yang|last4=Xia|first4=Xiao-luan|last5=Li|first5=Hai-fang|title=Brain functional network connectivity based on a visual task: visual information processing-related brain regions are significantly activated in the task state|journal=Neural Regeneration Research|date=2015|volume=10|issue=2|pages=298–307|doi=10.4103/1673-5374.152386|pmid=25883631|pmc=4392680 }}</ref><br />
<br />
<br />
*This network handles visual information processing.<br />
<br />
= = = = = = = = = = = = = 这个网络处理视觉信息。<br />
<br />
==Other networks==<br />
Different methods and data have identified several other brain networks, many of which greatly overlap or are subsets of more well-characterized core networks.<ref name="Uddin2019" /><br />
*Limbic<ref name="Bassett" /><ref name="Yeo" /><ref name="Bailey" /><br />
*Auditory<ref name="Yuan" /><ref name="Heine" /><br />
*Right/left executive<ref name="Yuan" /><ref name="Heine" /><br />
*Cerebellar<ref name="Bell" /><ref name="Heine" /><br />
*Spatial attention<ref name="Riedl" /><ref name="Bressler" /><br />
*Language<ref name="Bressler" /><ref name="Hutton" /><br />
* Lateral visual<ref name="Yuan" /><ref name="Bell" /><ref name="Heine" /><br />
*Temporal<ref name="Yeo" /><ref name="Shafiei" /><br />
*Visual perception/imagery<ref name="Hutton" /><br />
<br />
Different methods and data have identified several other brain networks, many of which greatly overlap or are subsets of more well-characterized core networks.<br />
*Limbic<br />
*Auditory<br />
*Right/left executive<br />
*Cerebellar<br />
*Spatial attention<br />
*Language<br />
*Lateral visual<br />
*Temporal<br />
*Visual perception/imagery<br />
<br />
其他网络不同的方法和数据已经确定了其他几个大脑网络，其中许多网络极大地重叠或者是更具特色的核心网络的子集。<br />
*边缘<br />
*听觉<br />
*右/左执行<br />
*小脑<br />
*空间注意<br />
*语言<br />
*外侧视觉<br />
*颞视知觉/图像<br />
<br />
==See also==<br />
*[[Complex network]]<br />
*[[Neural network]]<br />
<br />
*Complex network<br />
*Neural network<br />
<br />
=<nowiki>= = =</nowiki>= <br />
*复杂网络<br />
*<br />
*神经网络<br />
*<br />
<br />
== References==<br />
{{reflist|30em}}<br />
<br />
{{Human connectomics}}<br />
<br />
[[Category:Neuroscience]]<br />
[[Category:Neural coding]]<br />
[[Category:Neural circuits]]<br />
[[Category:Neurophysiology]]<br />
<br />
Category:Neuroscience<br />
Category:Neural coding<br />
Category:Neural circuits<br />
Category:Neurophysiology<br />
<br />
类别: 神经科学类别: 神经编码类别: 神经回路类别: 神经生理学<br />
<br />
<noinclude><br />
<br />
<small>This page was moved from [[wikipedia:en:Large-scale brain network]]. Its edit history can be viewed at [[大规模脑网络/edithistory]]</small></noinclude><br />
<br />
[[Category:待整理页面]]</div>Shenky20https://wiki.swarma.org/index.php?title=%E5%A4%A7%E8%A7%84%E6%A8%A1%E8%84%91%E7%BD%91%E7%BB%9C&diff=29981大规模脑网络2022-04-04T07:38:23Z<p>Shenky20：</p>
<hr />
<div>此词条由神经动力学模型读书会词条梳理志愿者Shenky20翻译审校，翻译字数共800，未经专家审核，带来阅读不便，请见谅。<br />
<br />
{{Short description|Collections of brain regions}}<br />
'''Large-scale brain networks''' (also known as '''intrinsic brain networks''') are collections of widespread [[brain regions]] showing [[Resting state fMRI#Functional|functional connectivity]] by statistical analysis of the [[Functional magnetic resonance imaging|fMRI]] [[BOLD signal]]<ref name=Riedl>{{cite journal|last1=Riedl|first1=Valentin|last2=Utz|first2=Lukas|last3=Castrillón|first3=Gabriel|last4=Grimmer|first4=Timo|last5=Rauschecker|first5=Josef P.|last6=Ploner|first6=Markus|last7=Friston|first7=Karl J.|last8=Drzezga|first8=Alexander|last9=Sorg|first9=Christian|title=Metabolic connectivity mapping reveals effective connectivity in the resting human brain|journal=PNAS|date=January 12, 2016|volume=113|issue=2|pages=428–433|doi=10.1073/pnas.1513752113|pmid=26712010|pmc=4720331|bibcode=2016PNAS..113..428R|doi-access=free}}</ref> or other recording methods such as [[Electroencephalography|EEG]],<ref>{{Cite journal|last1=Foster|first1=Brett L.|last2=Parvizi|first2=Josef|date=2012-03-01|title=Resting oscillations and cross-frequency coupling in the human posteromedial cortex|journal=NeuroImage|volume=60|issue=1|pages=384–391|doi=10.1016/j.neuroimage.2011.12.019|pmid=22227048|issn=1053-8119|pmc=3596417}}</ref> [[Positron emission tomography|PET]]<ref>{{Cite journal|last1=Buckner|first1=Randy L.|last2=Andrews‐Hanna|first2=Jessica R.|last3=Schacter|first3=Daniel L.|date=2008|title=The Brain's Default Network|journal=Annals of the New York Academy of Sciences|language=en|volume=1124|issue=1|pages=1–38|doi=10.1196/annals.1440.011|pmid=18400922|issn=1749-6632|bibcode=2008NYASA1124....1B|s2cid=3167595}}</ref> and [[Magnetoencephalography|MEG]].<ref>{{Cite journal|last1=Morris|first1=Peter G.|last2=Smith|first2=Stephen M.|last3=Barnes|first3=Gareth R.|last4=Stephenson|first4=Mary C.|last5=Hale|first5=Joanne R.|last6=Price|first6=Darren|last7=Luckhoo|first7=Henry|last8=Woolrich|first8=Mark|last9=Brookes|first9=Matthew J.|date=2011-10-04|title=Investigating the electrophysiological basis of resting state networks using magnetoencephalography|journal=Proceedings of the National Academy of Sciences|language=en|volume=108|issue=40|pages=16783–16788|doi=10.1073/pnas.1112685108|issn=0027-8424|pmid=21930901|pmc=3189080|bibcode=2011PNAS..10816783B|doi-access=free}}</ref> An emerging paradigm in neuroscience is that cognitive tasks are performed not by individual brain regions working in isolation but by networks consisting of several discrete brain regions that are said to be "functionally connected". Functional connectivity networks may be found using algorithms such as [[cluster analysis]], spatial [[independent component analysis]] (ICA), seed based, and others.<ref name="Petersen">{{cite journal|last1=Petersen|first1=Steven|last2=Sporns|first2=Olaf|title=Brain Networks and Cognitive Architectures|journal=Neuron|date=October 2015|volume=88|issue=1|pages=207–219|doi=10.1016/j.neuron.2015.09.027|pmid=26447582|pmc=4598639 }}</ref> Synchronized brain regions may also be identified using long-range synchronization of the EEG, MEG, or other dynamic brain signals.<ref name=Bressler>{{cite journal|last1=Bressler|first1=Steven L.|last2=Menon|first2=Vinod|s2cid=5967761|title=Large scale brain networks in cognition: emerging methods and principles|journal=Trends in Cognitive Sciences|date=June 2010|volume=14|issue=6|pages=233–290|doi=10.1016/j.tics.2010.04.004|url=http://www.cell.com/trends/cognitive-sciences/issue?pii=S1364-6613(10)X0005-5|accessdate=24 January 2016|pmid=20493761}}</ref><br />
<br />
<br />
Large-scale brain networks (also known as intrinsic brain networks) are collections of widespread brain regions showing functional connectivity by statistical analysis of the fMRI BOLD signal or other recording methods such as EEG, PET and MEG. An emerging paradigm in neuroscience is that cognitive tasks are performed not by individual brain regions working in isolation but by networks consisting of several discrete brain regions that are said to be "functionally connected". Functional connectivity networks may be found using algorithms such as cluster analysis, spatial independent component analysis (ICA), seed based, and others. Synchronized brain regions may also be identified using long-range synchronization of the EEG, MEG, or other dynamic brain signals.<br />
<br />
'''<font color="#ff8000">大规模脑网络Large-scale brain networks</font>'''(也称为'''<font color="#ff8000">内在大脑网络intrinsic brain networks</font>''')是在对'''<font color="#ff8000">基于血氧水平依赖效应的功能性磁共振成像信号fMRI BOLD signal</font>'''的统计分析或其他记录方法(如脑电图EEG、正电子发射断层扫描技术PET和脑磁图MEG)中，表现出功能连接的脑区的集合。根据神经科学中一个新出现的范式，认知任务不是由单个脑区独立执行的，而是由几个互不相连的脑区“功能连接”组成的网络执行的。功能连接网络可以通过数据聚类Cluster analysis、空间独立元素分析Spatial ICA、种子点方法seed-based等算法来发现。同步的脑区也可以用脑电图、脑磁图或其他动态脑信号的远程同步来识别。<br />
<br />
The set of identified brain areas that are linked together in a large-scale network varies with cognitive function.<ref name="Bressler2">{{cite journal|last1=Bressler|first1=Steven L.|title=Neurocognitive networks|journal=Scholarpedia|volume=3|issue=2|pages=1567|doi=10.4249/scholarpedia.1567|year=2008|bibcode=2008SchpJ...3.1567B|doi-access=free}}</ref> When the cognitive state is not explicit (i.e., the subject is at "rest"), the large-scale brain network is a [[Resting state fMRI|resting state]] network (RSN). As a physical system with graph-like properties,<ref name="Bressler" /> a large-scale brain network has both nodes and edges and cannot be identified simply by the co-activation of brain areas. In recent decades, the analysis of brain networks was made feasible by advances in imaging techniques as well as new tools from [[graph theory]] and [[Dynamical systems theory|dynamical systems]].<br />
<br />
The set of identified brain areas that are linked together in a large-scale network varies with cognitive function. When the cognitive state is not explicit (i.e., the subject is at "rest"), the large-scale brain network is a resting state network (RSN). As a physical system with graph-like properties, a large-scale brain network has both nodes and edges and cannot be identified simply by the co-activation of brain areas. In recent decades, the analysis of brain networks was made feasible by advances in imaging techniques as well as new tools from graph theory and dynamical systems.<br />
<br />
在一个大规模的网络中，连接在一起的大脑区域的集合因认知功能的不同而不同。当认知状态不明确(即主体处于“静止”状态)时，大规模的脑网络是一个静息状态网络(RSN)。作为一个具有图形特征的物理系统，大规模的脑网络既有节点又有边，不能简单地通过脑区的共同激活来识别。近几十年来，由于成像技术的进步以及图论和动力学系统的新工具，脑网络的分析变得可行。<br />
<br />
Large-scale brain networks are identified by their function and provide a coherent framework for understanding [[cognition]] by offering a neural model of how different cognitive functions emerge when different sets of brain regions join together as self-organized coalitions. The number and composition of the coalitions will vary with the algorithm and parameters used to identify them.<ref name="Yeo" /><ref>{{cite journal|last1=Abou Elseoud|first1=Ahmed|last2=Littow|first2=Harri|last3=Remes|first3=Jukka|last4=Starck|first4=Tuomo|last5=Nikkinen|first5=Juha|last6=Nissilä|first6=Juuso|last7=Timonen|first7=Markku|last8=Tervonen|first8=Osmo|last9=Kiviniemi1|first9=Vesa|title=Group-ICA Model Order Highlights Patterns of Functional Brain Connectivity|journal= Frontiers in Systems Neuroscience|date=2011-06-03|volume=5|pages=37|doi=10.3389/fnsys.2011.00037|pmid=21687724|pmc=3109774 |doi-access=free}}</ref> In one model, there is only the [[default mode network]] and the [[task-positive network]], but most current analyses show several networks, from a small handful to 17.<ref name="Yeo" /> The most common and stable networks are enumerated below. The regions participating in a functional network may be dynamically reconfigured.<ref name="Petersen" /><ref name="Bassett">{{cite journal|last1=Bassett|first1=Daniella|last2=Bertolero|first2=Max|title=How Matter Becomes Mind|journal=Scientific American|date=July 2019|volume=321|issue=1|page=32|url=https://www.scientificamerican.com/|accessdate=23 June 2019 }}</ref><br />
<br />
Large-scale brain networks are identified by their function and provide a coherent framework for understanding cognition by offering a neural model of how different cognitive functions emerge when different sets of brain regions join together as self-organized coalitions. The number and composition of the coalitions will vary with the algorithm and parameters used to identify them. In one model, there is only the default mode network and the task-positive network, but most current analyses show several networks, from a small handful to 17. The most common and stable networks are enumerated below. The regions participating in a functional network may be dynamically reconfigured.<br />
<br />
大规模的大脑网络是通过其功能识别的，并通过提供一个神经模型，说明当不同的大脑区域组合在一起形成自组织的联盟时，不同的认知功能是如何产生的，从而为理解认知提供了一个连贯的框架。联盟的数量和组成将根据识别联盟的算法和参数而有所不同。在一个模型中，只有默认模式网络和任务正向网络，但目前大多数分析显示，从少数到17个网络。下面列举了最常见和最稳定的网络。可以动态地重新配置参与功能网络的区域。<br />
<br />
Disruptions in activity in various networks have been implicated in neuropsychiatric disorders such as [[Depression (mood)|depression]], [[Alzheimer's disease|Alzheimer's]], [[Autism-spectrum disorder|autism spectrum disorder]], [[schizophrenia]], [[ADHD]]<ref>{{cite journal |last1=Griffiths |first1=Kristi R. |last2=Braund |first2=Taylor A. |last3=Kohn |first3=Michael R. |last4=Clarke |first4=Simon |last5=Williams |first5=Leanne M. |last6=Korgaonkar |first6=Mayuresh S. |title=Structural brain network topology underpinning ADHD and response to methylphenidate treatment |journal=Translational Psychiatry |date=2 March 2021 |volume=11 |issue=1 |pages=1–9 |doi=10.1038/s41398-021-01278-x | pmc=7925571 |pmid=33654073 |url=https://www.nature.com/articles/s41398-021-01278-x#citeas |access-date=16 November 2021}}</ref> and [[bipolar disorder]].<ref>{{Cite journal|url=https://www.researchgate.net/publication/51639686|title=Large-scale brain networks and psychopathology: A unifying triple network model|last=Menon|first=Vinod|s2cid=26653572|journal=Trends in Cognitive Sciences|date=2011-09-09|volume=15|issue=10|pages=483–506|doi=10.1016/j.tics.2011.08.003|pmid=21908230}}</ref><br />
<br />
Disruptions in activity in various networks have been implicated in neuropsychiatric disorders such as depression, Alzheimer's, autism spectrum disorder, schizophrenia, ADHD and bipolar disorder.<br />
<br />
各种网络活动的中断牵连到神经精神疾病，如抑郁症、老年痴呆症、自闭症光谱、精神分裂症、多动症和躁郁症。<br />
<br />
==Core networks==<br />
[[File:Heine2012x3010.png|thumb|An example that identified 10 large-scale brain networks from [[resting state fMRI]] activity through [[independent component analysis]].<ref name="Heine" />|链接=Special:FilePath/Heine2012x3010.png]]<br />
Because brain networks can be identified at various different resolutions and with various different neurobiological properties, there is no such thing as a universal atlas of brain networks that fits all circumstances.<ref>{{cite journal|last1=Eickhoff|first1=SB|last2=Yeo|first2=BTT|last3=Genon|first3=S|title=Imaging-based parcellations of the human brain.|journal=Nature Reviews. Neuroscience|date=November 2018|volume=19|issue=11|pages=672–686|doi=10.1038/s41583-018-0071-7|pmid=30305712|s2cid=52954265|url=http://juser.fz-juelich.de/record/856633/files/Eickhoff_Yeo_Genon_NRN_MainManuscriptInclFigures.pdf}}</ref> While acknowledging this problem, Uddin, Yeo, and Spreng proposed in 2019<ref name="Uddin2019">{{cite journal|last1=Uddin|first1=LQ|last2=Yeo|first2=BTT|last3=Spreng|first3=RN|title=Towards a Universal Taxonomy of Macro-scale Functional Human Brain Networks.|journal=Brain Topography|date=November 2019|volume=32|issue=6|pages=926–942|doi=10.1007/s10548-019-00744-6|pmid=31707621|pmc=7325607}}</ref> that the following six networks should be defined as core networks based on converging evidences from multiple studies<ref>{{cite journal|last1=Doucet|first1=GE|last2=Lee|first2=WH|last3=Frangou|first3=S|title=Evaluation of the spatial variability in the major resting-state networks across human brain functional atlases.|journal=Human Brain Mapping|date=2019-10-15|volume=40|issue=15|pages=4577–4587|doi=10.1002/hbm.24722|pmid=31322303|pmc=6771873}}</ref><ref name="Yeo" /><ref>{{cite journal|last1=Smith|first1=SM|last2=Fox|first2=PT|last3=Miller|first3=KL|last4=Glahn|first4=DC|last5=Fox|first5=PM|last6=Mackay|first6=CE|last7=Filippini|first7=N|last8=Watkins|first8=KE|last9=Toro|first9=R|last10=Laird|first10=AR|last11=Beckmann|first11=CF|title=Correspondence of the brain's functional architecture during activation and rest.|journal=Proceedings of the National Academy of Sciences of the United States of America|date=2009-08-04|volume=106|issue=31|pages=13040–5|doi=10.1073/pnas.0905267106|pmid=19620724|pmc=2722273|bibcode=2009PNAS..10613040S|doi-access=free}}</ref> to facilitate communication between researchers.<br />
<br />
<br />
Because brain networks can be identified at various different resolutions and with various different neurobiological properties, there is no such thing as a universal atlas of brain networks that fits all circumstances. While acknowledging this problem, Uddin, Yeo, and Spreng proposed in 2019 that the following six networks should be defined as core networks based on converging evidences from multiple studies to facilitate communication between researchers.<br />
<br />
因为大脑网络可以用不同的分辨率和不同的神经生物学特性来识别，所以没有适合所有情况的通用大脑网络图谱。在承认这个问题的同时，Uddin，Yeo，和 Spreng 在2019年提出，以下六个网络应该被定义为核心网络，基于来自多个研究的聚合证据，以促进研究人员之间的交流。<br />
<br />
=== Default Mode (Medial frontoparietal)===<br />
{{Main|Default mode network}}<br />
*The default mode network is active when an individual is awake and at rest. It preferentially activates when individuals focus on internally-oriented tasks such as daydreaming, envisioning the future, retrieving memories, and [[theory of mind]]. It is negatively correlated with brain systems that focus on external visual signals. It is the most widely researched network.<ref name="Bressler" /><ref name="Bassett" /><ref>{{Cite journal|date=2012-08-15|title=The serendipitous discovery of the brain's default network|journal=NeuroImage|language=en|volume=62|issue=2|pages=1137–1145|doi=10.1016/j.neuroimage.2011.10.035|pmid=22037421|issn=1053-8119|last1=Buckner|first1=Randy L.|s2cid=9880586}}</ref><ref name="Riedl" /><ref name="Yuan"><br />
{{cite journal|last1=Yuan|first1=Rui|last2=Di|first2=Xin|last3=Taylor|first3=Paul A.|last4=Gohel|first4=Suril|last5=Tsai|first5=Yuan-Hsiung|last6=Biswal|first6=Bharat B.|title=Functional topography of the thalamocortical system in human|journal=Brain Structure and Function|date=30 April 2015|doi=10.1007/s00429-015-1018-7|pmid=25924563|pmc=6363530|volume=221|issue=4|pages=1971–1984}}</ref><ref name="Bell">{{cite journal|last1=Bell|first1=Peter T.|last2=Shine|first2=James M.|title=Estimating Large-Scale Network Convergence in the Human Functional Connectome|journal=Brain Connectivity|date=2015-11-09|volume=5|issue=9|doi=10.1089/brain.2015.0348|pmid=26005099|pages=565–74}}</ref><ref name="Heine">{{cite journal|last1=Heine|first1=Lizette|last2=Soddu|first2=Andrea|last3=Gomez|first3=Francisco|last4=Vanhaudenhuyse|first4=Audrey|last5=Tshibanda|first5=Luaba|last6=Thonnard|first6=Marie|last7=Charland-Verville|first7=Vanessa|last8=Kirsch|first8=Murielle|last9=Laureys|first9=Steven|last10=Demertzi|first10=Athena|title=Resting state networks and consciousness. Alterations of multiple resting state network connectivity in physiological, pharmacological and pathological consciousness states.|journal=Frontiers in Psychology|date=2012|volume=3|pages=295|doi=10.3389/fpsyg.2012.00295|pmid=22969735|pmc=3427917|doi-access=free}}</ref><ref name="Yeo">{{cite journal|last1=Yeo|first1=B. T. Thomas|last2=Krienen|first2=Fenna M.|last3=Sepulcre|first3=Jorge|last4=Sabuncu|first4=Mert R.|last5=Lashkari|first5=Danial|last6=Hollinshead|first6=Marisa|last7=Roffman|first7=Joshua L.|last8=Smoller|first8=Jordan W.|last9=Zöllei|first9=Lilla|last10=Polimeni|first10=Jonathan R.|last11=Fischl|first11=Bruce|last12=Liu|first12=Hesheng|last13=Buckner|first13=Randy L.|title=The organization of the human cerebral cortex estimated by intrinsic functional connectivity|journal=Journal of Neurophysiology|date=2011-09-01|volume=106|issue=3|pages=1125–1165|doi=10.1152/jn.00338.2011|pmid=21653723|pmc=3174820|bibcode=2011NatSD...2E0031H }}</ref><ref name="Shafiei">{{cite journal|last1=Shafiei|first1=Golia|last2=Zeighami|first2=Yashar|last3=Clark|first3=Crystal A.|last4=Coull|first4=Jennifer T.|last5=Nagano-Saito|first5=Atsuko|last6=Leyton|first6=Marco|last7=Dagher|first7=Alain|last8=Mišić|first8=Bratislav|title=Dopamine Signaling Modulates the Stability and Integration of Intrinsic Brain Networks|journal=Cerebral Cortex|date=2018-10-01|volume=29|issue=1|pages=397–409|doi=10.1093/cercor/bhy264|pmid=30357316|pmc=6294404 }}</ref><ref name="Bailey">{{cite journal|last1=Bailey|first1=Stephen K.|last2=Aboud|first2=Katherine S.|last3=Nguyen|first3=Tin Q.|last4=Cutting|first4=Laurie E.|title=Applying a network framework to the neurobiology of reading and dyslexia|journal=Journal of Neurodevelopmental Disorders|date=13 December 2018|volume=10|issue=1|page=37|doi=10.1186/s11689-018-9251-z|pmid=30541433|pmc=6291929 }}</ref><br />
<br />
<br />
*The default mode network is active when an individual is awake and at rest. It preferentially activates when individuals focus on internally-oriented tasks such as daydreaming, envisioning the future, retrieving memories, and theory of mind. It is negatively correlated with brain systems that focus on external visual signals. It is the most widely researched network.<br />
<br />
<br />
=<nowiki>= = 默认模式(Medial frontoparietal) = =</nowiki>= <br />
*默认模式网络在个人清醒和休息时处于活动状态。当个体专注于面向内部的任务时，比如做白日梦、展望未来、回忆和心理理论，它就会优先激活。它与专注于外部视觉信号的大脑系统负相关。它是研究最广泛的网络。<br />
<br />
=== Salience (Midcingulo-Insular)===<br />
{{Main|Salience network}}<br />
*The salience network consists of several structures, including the anterior (bilateral) insula, dorsal anterior cingulate cortex, and three subcortical structures which are the ventral striatum, substantia nigra/ventral tegmental region.<ref>{{Cite journal|last1=Steimke|first1=Rosa|last2=Nomi|first2=Jason S.|last3=Calhoun|first3=Vince D.|last4=Stelzel|first4=Christine|last5=Paschke|first5=Lena M.|last6=Gaschler|first6=Robert|last7=Goschke|first7=Thomas|last8=Walter|first8=Henrik|last9=Uddin|first9=Lucina Q.|date=2017-12-01|title=Salience network dynamics underlying successful resistance of temptation|journal=Social Cognitive and Affective Neuroscience|language=en|volume=12|issue=12|pages=1928–1939|doi=10.1093/scan/nsx123|pmid=29048582|pmc=5716209|issn=1749-5016|doi-access=free}}</ref><ref name=":0">{{Citation|last=Menon|first=V.|title=Brain Mapping|chapter=Salience Network|date=2015-01-01|chapter-url=http://www.sciencedirect.com/science/article/pii/B978012397025100052X|pages=597–611|editor-last=Toga|editor-first=Arthur W.|publisher=Academic Press|isbn=978-0-12-397316-0|access-date=2019-12-08|doi=10.1016/B978-0-12-397025-1.00052-X}}</ref> It plays the key role of monitoring the [[Salience (neuroscience)|salience]] of external inputs and internal brain events.<ref name="Riedl" /><ref name="Bressler" /><ref name="Bassett" /><ref name="Yuan" /><ref name="Heine" /><ref name="Yeo" /><ref name="Shafiei" /> Specifically, it aids in directing attention by identifying important biological and cognitive events.<ref name=":0" /><ref name="Bailey" /><br />
*This network includes the ventral attention network, which primarily includes the [[temporoparietal junction]] and the ventral [[frontal cortex]] of the right hemisphere.<ref name="Uddin2019" /><ref name="Vossel" /> These areas respond when behaviorally relevant stimuli occur unexpectedly.<ref name="Vossel" /> The ventral attention network is inhibited during focused attention in which top-down processing is being used, such as when visually searching for something. This response may prevent goal-driven attention from being distracted by non-relevant stimuli. It becomes active again when the target or relevant information about the target is found.<ref name="Vossel" /><ref>{{Cite journal|last1=Shulman|first1=Gordon L.|last2=McAvoy|first2=Mark P.|last3=Cowan|first3=Melanie C.|last4=Astafiev|first4=Serguei V.|last5=Tansy|first5=Aaron P.|last6=d'Avossa|first6=Giovanni|last7=Corbetta|first7=Maurizio|date=2003-11-01|title=Quantitative Analysis of Attention and Detection Signals During Visual Search|journal=Journal of Neurophysiology|volume=90|issue=5|pages=3384–3397|doi=10.1152/jn.00343.2003|pmid=12917383|issn=0022-3077}}</ref><br />
<br />
<br />
*The salience network consists of several structures, including the anterior (bilateral) insula, dorsal anterior cingulate cortex, and three subcortical structures which are the ventral striatum, substantia nigra/ventral tegmental region. It plays the key role of monitoring the salience of external inputs and internal brain events. Specifically, it aids in directing attention by identifying important biological and cognitive events.<br />
*This network includes the ventral attention network, which primarily includes the temporoparietal junction and the ventral frontal cortex of the right hemisphere. These areas respond when behaviorally relevant stimuli occur unexpectedly. The ventral attention network is inhibited during focused attention in which top-down processing is being used, such as when visually searching for something. This response may prevent goal-driven attention from being distracted by non-relevant stimuli. It becomes active again when the target or relevant information about the target is found.<br />
<br />
突显网络由前(双)岛、前扣带回皮层和腹侧纹状体、黑质/腹侧被盖区3个皮层下结构组成。它起着监测外部输入和内部脑事件的显著性的关键作用。具体来说，它通过识别重要的生物学和认知事件来帮助引导注意力。<br />
*这个网络包括腹侧注意网络，主要包括右半球的颞顶联合区和腹侧额叶皮层。当行为相关的刺激意外发生时，这些区域会做出反应。腹侧注意网络在使用自上而下加工的集中注意过程中被抑制，例如在视觉搜索某物时。这种反应可以防止目标驱动的注意力被非相关的刺激分散。当找到目标或关于目标的相关信息时，它再次激活。<br />
<br />
===Attention (Dorsal frontoparietal)===<br />
{{Main|Dorsal attention network}}<br />
* This network is involved in the voluntary, top-down deployment of attention.<ref name="Riedl" /><ref name="Yuan" /><ref name="Bell" /><ref name="Yeo" /><ref name="Shafiei" /><ref name="Vossel">{{cite journal|last1=Vossel|first1=Simone|last2=Geng|first2=Joy J.|last3=Fink|first3=Gereon R.|title=Dorsal and Ventral Attention Systems: Distinct Neural Circuits but Collaborative Roles|journal=The Neuroscientist|date=2014|volume=20|issue=2|pages=150–159|doi=10.1177/1073858413494269|pmid=23835449|pmc=4107817}}</ref><ref name="Hutton">{{cite journal|last1=Hutton|first1=John S.|last2=Dudley|first2=Jonathan|last3=Horowitz-Kraus|first3=Tzipi|last4=DeWitt|first4=Tom|last5=Holland|first5=Scott K.|title=Functional Connectivity of Attention, Visual, and Language Networks During Audio, Illustrated, and Animated Stories in Preschool-Age Children|journal=Brain Connectivity|date=1 September 2019|volume=9|issue=7|pages=580–592|doi=10.1089/brain.2019.0679|pmid=31144523|pmc=6775495|ref=Hutton}}</ref> Within the dorsal attention network, the intraparietal sulcus and frontal eye fields influence the visual areas of the brain. These influencing factors allow for the orientation of attention.<ref>{{Cite journal|last1=Fox|first1=Michael D.|last2=Corbetta|first2=Maurizio|last3=Snyder|first3=Abraham Z.|last4=Vincent|first4=Justin L.|last5=Raichle|first5=Marcus E.|date=2006-06-27|title=Spontaneous neuronal activity distinguishes human dorsal and ventral attention systems|journal=Proceedings of the National Academy of Sciences|language=en|volume=103|issue=26|pages=10046–10051|doi=10.1073/pnas.0604187103|issn=0027-8424|pmid=16788060|pmc=1480402|bibcode=2006PNAS..10310046F|doi-access=free}}</ref><ref name="Vossel" /><ref name="Bailey" /><br />
<br />
<br />
*This network is involved in the voluntary, top-down deployment of attention. Within the dorsal attention network, the intraparietal sulcus and frontal eye fields influence the visual areas of the brain. These influencing factors allow for the orientation of attention.<br />
<br />
=<nowiki>= = 注意力(背侧额顶骨) = =</nowiki>= <br />
*这个网络参与了自发的、自上而下的注意力分配。在背侧注意网络中，顶内沟和额眼影响大脑的视觉区域。这些影响因素决定了注意力的方向。<br />
<br />
===Control (Lateral frontoparietal)===<br />
{{Main|Frontoparietal network}}<br />
*This network initiates and modulates cognitive control and comprises 18 sub-regions of the brain.<ref>{{Cite journal|last1=Scolari|first1=Miranda|last2=Seidl-Rathkopf|first2=Katharina N|last3=Kastner|first3=Sabine|date=2015-02-01|title=Functions of the human frontoparietal attention network: Evidence from neuroimaging|journal=Current Opinion in Behavioral Sciences|series=Cognitive control|volume=1|pages=32–39|doi=10.1016/j.cobeha.2014.08.003|issn=2352-1546|pmid=27398396|pmc=4936532}}</ref> There is a strong correlation between fluid intelligence and the involvement of the fronto-parietal network with other networks.<ref>{{Cite journal|last1=Marek|first1=Scott|last2=Dosenbach|first2=Nico U. F.|date=June 2018|title=The frontoparietal network: function, electrophysiology, and importance of individual precision mapping|journal=Dialogues in Clinical Neuroscience|volume=20|issue=2|pages=133–140|doi=10.31887/DCNS.2018.20.2/smarek|issn=1294-8322|pmc=6136121|pmid=30250390}}</ref><br />
<br />
<br />
*This network initiates and modulates cognitive control and comprises 18 sub-regions of the brain. There is a strong correlation between fluid intelligence and the involvement of the fronto-parietal network with other networks.<br />
<br />
=<nowiki>= = 控制(侧额顶骨) = =</nowiki>= <br />
*这个网络启动和调节认知控制，包括大脑的18个亚区。在流体智力和额顶网络与其他网络的参与之间有很强的相关性。<br />
<br />
*Versions of this network have also been called the central executive (or executive control) network and the cognitive control network.<ref name="Uddin2019" /><br />
<br />
*Versions of this network have also been called the central executive (or executive control) network and the cognitive control network.<br />
<br />
<br />
*这种网络的版本也被称为中央执行(或执行控制)网络和认知控制网络。<br />
<br />
===Sensorimotor or Somatomotor (Pericentral)===<br />
{{Main|Sensorimotor network}}<br />
*This network processes somatosensory information and coordinates motion.<ref name="Heine" /><ref name="Yeo" /><ref name="Shafiei" /><ref name="Bassett" /><ref name="Yuan" /> The [[auditory cortex]] may be included.<ref name="Uddin2019" /><ref name="Yeo" /><br />
<br />
<br />
*This network processes somatosensory information and coordinates motion. The auditory cortex may be included.<br />
<br />
===Sensorimotor or Somatomotor (Pericentral)===<br />
<br />
<br />
*This network processes somatosensory information and coordinates motion.听觉皮层可能也包括在内。<br />
<br />
===Visual (Occipital)===<br />
{{See|Visual cortex}}<br />
*This network handles visual information processing.<ref name="Yang">{{cite journal|last1=Yang|first1=Yan-li|last2=Deng|first2=Hong-xia|last3=Xing|first3=Gui-yang|last4=Xia|first4=Xiao-luan|last5=Li|first5=Hai-fang|title=Brain functional network connectivity based on a visual task: visual information processing-related brain regions are significantly activated in the task state|journal=Neural Regeneration Research|date=2015|volume=10|issue=2|pages=298–307|doi=10.4103/1673-5374.152386|pmid=25883631|pmc=4392680 }}</ref><br />
<br />
<br />
*This network handles visual information processing.<br />
<br />
= = = = = = = = = = = = = 这个网络处理视觉信息。<br />
<br />
==Other networks==<br />
Different methods and data have identified several other brain networks, many of which greatly overlap or are subsets of more well-characterized core networks.<ref name="Uddin2019" /><br />
*Limbic<ref name="Bassett" /><ref name="Yeo" /><ref name="Bailey" /><br />
*Auditory<ref name="Yuan" /><ref name="Heine" /><br />
*Right/left executive<ref name="Yuan" /><ref name="Heine" /><br />
*Cerebellar<ref name="Bell" /><ref name="Heine" /><br />
*Spatial attention<ref name="Riedl" /><ref name="Bressler" /><br />
*Language<ref name="Bressler" /><ref name="Hutton" /><br />
*Lateral visual<ref name="Yuan" /><ref name="Bell" /><ref name="Heine" /><br />
* Temporal<ref name="Yeo" /><ref name="Shafiei" /><br />
*Visual perception/imagery<ref name="Hutton" /><br />
<br />
Different methods and data have identified several other brain networks, many of which greatly overlap or are subsets of more well-characterized core networks.<br />
*Limbic<br />
*Auditory<br />
*Right/left executive<br />
*Cerebellar<br />
*Spatial attention<br />
*Language<br />
*Lateral visual<br />
*Temporal<br />
*Visual perception/imagery<br />
<br />
其他网络不同的方法和数据已经确定了其他几个大脑网络，其中许多网络极大地重叠或者是更具特色的核心网络的子集。<br />
*边缘<br />
*听觉<br />
*右/左执行<br />
*小脑<br />
*空间注意<br />
*语言<br />
*外侧视觉<br />
*颞视知觉/图像<br />
<br />
==See also==<br />
*[[Complex network]]<br />
*[[Neural network]]<br />
<br />
*Complex network<br />
*Neural network<br />
<br />
=<nowiki>= = =</nowiki>= <br />
*复杂网络<br />
*<br />
*神经网络<br />
*<br />
<br />
==References==<br />
{{reflist|30em}}<br />
<br />
{{Human connectomics}}<br />
<br />
[[Category:Neuroscience]]<br />
[[Category:Neural coding]]<br />
[[Category:Neural circuits]]<br />
[[Category:Neurophysiology]]<br />
<br />
Category:Neuroscience<br />
Category:Neural coding<br />
Category:Neural circuits<br />
Category:Neurophysiology<br />
<br />
类别: 神经科学类别: 神经编码类别: 神经回路类别: 神经生理学<br />
<br />
<noinclude><br />
<br />
<small>This page was moved from [[wikipedia:en:Large-scale brain network]]. Its edit history can be viewed at [[大规模脑网络/edithistory]]</small></noinclude><br />
<br />
[[Category:待整理页面]]</div>Shenky20https://wiki.swarma.org/index.php?title=%E5%A4%A7%E8%A7%84%E6%A8%A1%E8%84%91%E7%BD%91%E7%BB%9C&diff=29980大规模脑网络2022-04-04T07:29:12Z<p>Shenky20：</p>
<hr />
<div>此词条由神经动力学模型读书会词条梳理志愿者Shenky20翻译审校，翻译字数共800，未经专家审核，带来阅读不便，请见谅。<br />
<br />
{{Short description|Collections of brain regions}}<br />
'''Large-scale brain networks''' (also known as '''intrinsic brain networks''') are collections of widespread [[brain regions]] showing [[Resting state fMRI#Functional|functional connectivity]] by statistical analysis of the [[Functional magnetic resonance imaging|fMRI]] [[BOLD signal]]<ref name=Riedl>{{cite journal|last1=Riedl|first1=Valentin|last2=Utz|first2=Lukas|last3=Castrillón|first3=Gabriel|last4=Grimmer|first4=Timo|last5=Rauschecker|first5=Josef P.|last6=Ploner|first6=Markus|last7=Friston|first7=Karl J.|last8=Drzezga|first8=Alexander|last9=Sorg|first9=Christian|title=Metabolic connectivity mapping reveals effective connectivity in the resting human brain|journal=PNAS|date=January 12, 2016|volume=113|issue=2|pages=428–433|doi=10.1073/pnas.1513752113|pmid=26712010|pmc=4720331|bibcode=2016PNAS..113..428R|doi-access=free}}</ref> or other recording methods such as [[Electroencephalography|EEG]],<ref>{{Cite journal|last1=Foster|first1=Brett L.|last2=Parvizi|first2=Josef|date=2012-03-01|title=Resting oscillations and cross-frequency coupling in the human posteromedial cortex|journal=NeuroImage|volume=60|issue=1|pages=384–391|doi=10.1016/j.neuroimage.2011.12.019|pmid=22227048|issn=1053-8119|pmc=3596417}}</ref> [[Positron emission tomography|PET]]<ref>{{Cite journal|last1=Buckner|first1=Randy L.|last2=Andrews‐Hanna|first2=Jessica R.|last3=Schacter|first3=Daniel L.|date=2008|title=The Brain's Default Network|journal=Annals of the New York Academy of Sciences|language=en|volume=1124|issue=1|pages=1–38|doi=10.1196/annals.1440.011|pmid=18400922|issn=1749-6632|bibcode=2008NYASA1124....1B|s2cid=3167595}}</ref> and [[Magnetoencephalography|MEG]].<ref>{{Cite journal|last1=Morris|first1=Peter G.|last2=Smith|first2=Stephen M.|last3=Barnes|first3=Gareth R.|last4=Stephenson|first4=Mary C.|last5=Hale|first5=Joanne R.|last6=Price|first6=Darren|last7=Luckhoo|first7=Henry|last8=Woolrich|first8=Mark|last9=Brookes|first9=Matthew J.|date=2011-10-04|title=Investigating the electrophysiological basis of resting state networks using magnetoencephalography|journal=Proceedings of the National Academy of Sciences|language=en|volume=108|issue=40|pages=16783–16788|doi=10.1073/pnas.1112685108|issn=0027-8424|pmid=21930901|pmc=3189080|bibcode=2011PNAS..10816783B|doi-access=free}}</ref> An emerging paradigm in neuroscience is that cognitive tasks are performed not by individual brain regions working in isolation but by networks consisting of several discrete brain regions that are said to be "functionally connected". Functional connectivity networks may be found using algorithms such as [[cluster analysis]], spatial [[independent component analysis]] (ICA), seed based, and others.<ref name="Petersen">{{cite journal|last1=Petersen|first1=Steven|last2=Sporns|first2=Olaf|title=Brain Networks and Cognitive Architectures|journal=Neuron|date=October 2015|volume=88|issue=1|pages=207–219|doi=10.1016/j.neuron.2015.09.027|pmid=26447582|pmc=4598639 }}</ref> Synchronized brain regions may also be identified using long-range synchronization of the EEG, MEG, or other dynamic brain signals.<ref name=Bressler>{{cite journal|last1=Bressler|first1=Steven L.|last2=Menon|first2=Vinod|s2cid=5967761|title=Large scale brain networks in cognition: emerging methods and principles|journal=Trends in Cognitive Sciences|date=June 2010|volume=14|issue=6|pages=233–290|doi=10.1016/j.tics.2010.04.004|url=http://www.cell.com/trends/cognitive-sciences/issue?pii=S1364-6613(10)X0005-5|accessdate=24 January 2016|pmid=20493761}}</ref><br />
<br />
<br />
Large-scale brain networks (also known as intrinsic brain networks) are collections of widespread brain regions showing functional connectivity by statistical analysis of the fMRI BOLD signal or other recording methods such as EEG, PET and MEG. An emerging paradigm in neuroscience is that cognitive tasks are performed not by individual brain regions working in isolation but by networks consisting of several discrete brain regions that are said to be "functionally connected". Functional connectivity networks may be found using algorithms such as cluster analysis, spatial independent component analysis (ICA), seed based, and others. Synchronized brain regions may also be identified using long-range synchronization of the EEG, MEG, or other dynamic brain signals.<br />
<br />
大规模脑网络Large-scale brain networks(也称为内在大脑网络intrinsic brain networks)是在对基于血氧水平依赖BOLD效应的功能性磁共振成像fMRI信号的统计分析或其他记录方法(如脑电图EEG、正电子发射断层扫描技术PET和脑磁图MEG)中，表现出功能连接的脑区的集合。根据神经科学中一个新出现的范式，认知任务不是由单个脑区独立执行的，而是由几个互不相连的脑区“功能连接”组成的网络执行的。功能连接网络可以通过数据聚类Cluster analysis、空间独立元素分析Spatial ICA、种子点方法seed-based等算法来发现。同步的脑区也可以用脑电图、脑磁图或其他动态脑信号的远程同步来识别。<br />
<br />
The set of identified brain areas that are linked together in a large-scale network varies with cognitive function.<ref name="Bressler2">{{cite journal|last1=Bressler|first1=Steven L.|title=Neurocognitive networks|journal=Scholarpedia|volume=3|issue=2|pages=1567|doi=10.4249/scholarpedia.1567|year=2008|bibcode=2008SchpJ...3.1567B|doi-access=free}}</ref> When the cognitive state is not explicit (i.e., the subject is at "rest"), the large-scale brain network is a [[Resting state fMRI|resting state]] network (RSN). As a physical system with graph-like properties,<ref name="Bressler" /> a large-scale brain network has both nodes and edges and cannot be identified simply by the co-activation of brain areas. In recent decades, the analysis of brain networks was made feasible by advances in imaging techniques as well as new tools from [[graph theory]] and [[Dynamical systems theory|dynamical systems]].<br />
<br />
The set of identified brain areas that are linked together in a large-scale network varies with cognitive function. When the cognitive state is not explicit (i.e., the subject is at "rest"), the large-scale brain network is a resting state network (RSN). As a physical system with graph-like properties, a large-scale brain network has both nodes and edges and cannot be identified simply by the co-activation of brain areas. In recent decades, the analysis of brain networks was made feasible by advances in imaging techniques as well as new tools from graph theory and dynamical systems.<br />
<br />
在一个大规模的网络中，连接在一起的大脑区域的集合因认知功能的不同而不同。当认知状态不明确(即主体处于“静止”状态)时，大规模的脑网络是一个静息状态网络(RSN)。作为一个具有图形特征的物理系统，大规模的脑网络既有节点又有边，不能简单地通过脑区的共同激活来识别。近几十年来，由于成像技术的进步以及图论和动力学系统的新工具，脑网络的分析变得可行。<br />
<br />
Large-scale brain networks are identified by their function and provide a coherent framework for understanding [[cognition]] by offering a neural model of how different cognitive functions emerge when different sets of brain regions join together as self-organized coalitions. The number and composition of the coalitions will vary with the algorithm and parameters used to identify them.<ref name="Yeo" /><ref>{{cite journal|last1=Abou Elseoud|first1=Ahmed|last2=Littow|first2=Harri|last3=Remes|first3=Jukka|last4=Starck|first4=Tuomo|last5=Nikkinen|first5=Juha|last6=Nissilä|first6=Juuso|last7=Timonen|first7=Markku|last8=Tervonen|first8=Osmo|last9=Kiviniemi1|first9=Vesa|title=Group-ICA Model Order Highlights Patterns of Functional Brain Connectivity|journal= Frontiers in Systems Neuroscience|date=2011-06-03|volume=5|pages=37|doi=10.3389/fnsys.2011.00037|pmid=21687724|pmc=3109774 |doi-access=free}}</ref> In one model, there is only the [[default mode network]] and the [[task-positive network]], but most current analyses show several networks, from a small handful to 17.<ref name="Yeo" /> The most common and stable networks are enumerated below. The regions participating in a functional network may be dynamically reconfigured.<ref name="Petersen" /><ref name="Bassett">{{cite journal|last1=Bassett|first1=Daniella|last2=Bertolero|first2=Max|title=How Matter Becomes Mind|journal=Scientific American|date=July 2019|volume=321|issue=1|page=32|url=https://www.scientificamerican.com/|accessdate=23 June 2019 }}</ref><br />
<br />
Large-scale brain networks are identified by their function and provide a coherent framework for understanding cognition by offering a neural model of how different cognitive functions emerge when different sets of brain regions join together as self-organized coalitions. The number and composition of the coalitions will vary with the algorithm and parameters used to identify them. In one model, there is only the default mode network and the task-positive network, but most current analyses show several networks, from a small handful to 17. The most common and stable networks are enumerated below. The regions participating in a functional network may be dynamically reconfigured.<br />
<br />
大规模的大脑网络是通过其功能识别的，并通过提供一个神经模型，说明当不同的大脑区域组合在一起形成自组织的联盟时，不同的认知功能是如何产生的，从而为理解认知提供了一个连贯的框架。联盟的数量和组成将根据识别联盟的算法和参数而有所不同。在一个模型中，只有默认模式网络和任务正向网络，但目前大多数分析显示，从少数到17个网络。下面列举了最常见和最稳定的网络。可以动态地重新配置参与功能网络的区域。<br />
<br />
Disruptions in activity in various networks have been implicated in neuropsychiatric disorders such as [[Depression (mood)|depression]], [[Alzheimer's disease|Alzheimer's]], [[Autism-spectrum disorder|autism spectrum disorder]], [[schizophrenia]], [[ADHD]]<ref>{{cite journal |last1=Griffiths |first1=Kristi R. |last2=Braund |first2=Taylor A. |last3=Kohn |first3=Michael R. |last4=Clarke |first4=Simon |last5=Williams |first5=Leanne M. |last6=Korgaonkar |first6=Mayuresh S. |title=Structural brain network topology underpinning ADHD and response to methylphenidate treatment |journal=Translational Psychiatry |date=2 March 2021 |volume=11 |issue=1 |pages=1–9 |doi=10.1038/s41398-021-01278-x | pmc=7925571 |pmid=33654073 |url=https://www.nature.com/articles/s41398-021-01278-x#citeas |access-date=16 November 2021}}</ref> and [[bipolar disorder]].<ref>{{Cite journal|url=https://www.researchgate.net/publication/51639686|title=Large-scale brain networks and psychopathology: A unifying triple network model|last=Menon|first=Vinod|s2cid=26653572|journal=Trends in Cognitive Sciences|date=2011-09-09|volume=15|issue=10|pages=483–506|doi=10.1016/j.tics.2011.08.003|pmid=21908230}}</ref><br />
<br />
Disruptions in activity in various networks have been implicated in neuropsychiatric disorders such as depression, Alzheimer's, autism spectrum disorder, schizophrenia, ADHD and bipolar disorder.<br />
<br />
各种网络活动的中断牵连到神经精神疾病，如抑郁症、老年痴呆症、自闭症光谱、精神分裂症、多动症和躁郁症。<br />
<br />
==Core networks==<br />
[[File:Heine2012x3010.png|thumb|An example that identified 10 large-scale brain networks from [[resting state fMRI]] activity through [[independent component analysis]].<ref name="Heine" />|链接=Special:FilePath/Heine2012x3010.png]]<br />
Because brain networks can be identified at various different resolutions and with various different neurobiological properties, there is no such thing as a universal atlas of brain networks that fits all circumstances.<ref>{{cite journal|last1=Eickhoff|first1=SB|last2=Yeo|first2=BTT|last3=Genon|first3=S|title=Imaging-based parcellations of the human brain.|journal=Nature Reviews. Neuroscience|date=November 2018|volume=19|issue=11|pages=672–686|doi=10.1038/s41583-018-0071-7|pmid=30305712|s2cid=52954265|url=http://juser.fz-juelich.de/record/856633/files/Eickhoff_Yeo_Genon_NRN_MainManuscriptInclFigures.pdf}}</ref> While acknowledging this problem, Uddin, Yeo, and Spreng proposed in 2019<ref name="Uddin2019">{{cite journal|last1=Uddin|first1=LQ|last2=Yeo|first2=BTT|last3=Spreng|first3=RN|title=Towards a Universal Taxonomy of Macro-scale Functional Human Brain Networks.|journal=Brain Topography|date=November 2019|volume=32|issue=6|pages=926–942|doi=10.1007/s10548-019-00744-6|pmid=31707621|pmc=7325607}}</ref> that the following six networks should be defined as core networks based on converging evidences from multiple studies<ref>{{cite journal|last1=Doucet|first1=GE|last2=Lee|first2=WH|last3=Frangou|first3=S|title=Evaluation of the spatial variability in the major resting-state networks across human brain functional atlases.|journal=Human Brain Mapping|date=2019-10-15|volume=40|issue=15|pages=4577–4587|doi=10.1002/hbm.24722|pmid=31322303|pmc=6771873}}</ref><ref name="Yeo" /><ref>{{cite journal|last1=Smith|first1=SM|last2=Fox|first2=PT|last3=Miller|first3=KL|last4=Glahn|first4=DC|last5=Fox|first5=PM|last6=Mackay|first6=CE|last7=Filippini|first7=N|last8=Watkins|first8=KE|last9=Toro|first9=R|last10=Laird|first10=AR|last11=Beckmann|first11=CF|title=Correspondence of the brain's functional architecture during activation and rest.|journal=Proceedings of the National Academy of Sciences of the United States of America|date=2009-08-04|volume=106|issue=31|pages=13040–5|doi=10.1073/pnas.0905267106|pmid=19620724|pmc=2722273|bibcode=2009PNAS..10613040S|doi-access=free}}</ref> to facilitate communication between researchers.<br />
<br />
<br />
Because brain networks can be identified at various different resolutions and with various different neurobiological properties, there is no such thing as a universal atlas of brain networks that fits all circumstances. While acknowledging this problem, Uddin, Yeo, and Spreng proposed in 2019 that the following six networks should be defined as core networks based on converging evidences from multiple studies to facilitate communication between researchers.<br />
<br />
因为大脑网络可以用不同的分辨率和不同的神经生物学特性来识别，所以没有适合所有情况的通用大脑网络图谱。在承认这个问题的同时，Uddin，Yeo，和 Spreng 在2019年提出，以下六个网络应该被定义为核心网络，基于来自多个研究的聚合证据，以促进研究人员之间的交流。<br />
<br />
===Default Mode (Medial frontoparietal)===<br />
{{Main|Default mode network}}<br />
* The default mode network is active when an individual is awake and at rest. It preferentially activates when individuals focus on internally-oriented tasks such as daydreaming, envisioning the future, retrieving memories, and [[theory of mind]]. It is negatively correlated with brain systems that focus on external visual signals. It is the most widely researched network.<ref name="Bressler" /><ref name="Bassett" /><ref>{{Cite journal|date=2012-08-15|title=The serendipitous discovery of the brain's default network|journal=NeuroImage|language=en|volume=62|issue=2|pages=1137–1145|doi=10.1016/j.neuroimage.2011.10.035|pmid=22037421|issn=1053-8119|last1=Buckner|first1=Randy L.|s2cid=9880586}}</ref><ref name="Riedl" /><ref name="Yuan"><br />
{{cite journal|last1=Yuan|first1=Rui|last2=Di|first2=Xin|last3=Taylor|first3=Paul A.|last4=Gohel|first4=Suril|last5=Tsai|first5=Yuan-Hsiung|last6=Biswal|first6=Bharat B.|title=Functional topography of the thalamocortical system in human|journal=Brain Structure and Function|date=30 April 2015|doi=10.1007/s00429-015-1018-7|pmid=25924563|pmc=6363530|volume=221|issue=4|pages=1971–1984}}</ref><ref name="Bell">{{cite journal|last1=Bell|first1=Peter T.|last2=Shine|first2=James M.|title=Estimating Large-Scale Network Convergence in the Human Functional Connectome|journal=Brain Connectivity|date=2015-11-09|volume=5|issue=9|doi=10.1089/brain.2015.0348|pmid=26005099|pages=565–74}}</ref><ref name="Heine">{{cite journal|last1=Heine|first1=Lizette|last2=Soddu|first2=Andrea|last3=Gomez|first3=Francisco|last4=Vanhaudenhuyse|first4=Audrey|last5=Tshibanda|first5=Luaba|last6=Thonnard|first6=Marie|last7=Charland-Verville|first7=Vanessa|last8=Kirsch|first8=Murielle|last9=Laureys|first9=Steven|last10=Demertzi|first10=Athena|title=Resting state networks and consciousness. Alterations of multiple resting state network connectivity in physiological, pharmacological and pathological consciousness states.|journal=Frontiers in Psychology|date=2012|volume=3|pages=295|doi=10.3389/fpsyg.2012.00295|pmid=22969735|pmc=3427917|doi-access=free}}</ref><ref name="Yeo">{{cite journal|last1=Yeo|first1=B. T. Thomas|last2=Krienen|first2=Fenna M.|last3=Sepulcre|first3=Jorge|last4=Sabuncu|first4=Mert R.|last5=Lashkari|first5=Danial|last6=Hollinshead|first6=Marisa|last7=Roffman|first7=Joshua L.|last8=Smoller|first8=Jordan W.|last9=Zöllei|first9=Lilla|last10=Polimeni|first10=Jonathan R.|last11=Fischl|first11=Bruce|last12=Liu|first12=Hesheng|last13=Buckner|first13=Randy L.|title=The organization of the human cerebral cortex estimated by intrinsic functional connectivity|journal=Journal of Neurophysiology|date=2011-09-01|volume=106|issue=3|pages=1125–1165|doi=10.1152/jn.00338.2011|pmid=21653723|pmc=3174820|bibcode=2011NatSD...2E0031H }}</ref><ref name="Shafiei">{{cite journal|last1=Shafiei|first1=Golia|last2=Zeighami|first2=Yashar|last3=Clark|first3=Crystal A.|last4=Coull|first4=Jennifer T.|last5=Nagano-Saito|first5=Atsuko|last6=Leyton|first6=Marco|last7=Dagher|first7=Alain|last8=Mišić|first8=Bratislav|title=Dopamine Signaling Modulates the Stability and Integration of Intrinsic Brain Networks|journal=Cerebral Cortex|date=2018-10-01|volume=29|issue=1|pages=397–409|doi=10.1093/cercor/bhy264|pmid=30357316|pmc=6294404 }}</ref><ref name="Bailey">{{cite journal|last1=Bailey|first1=Stephen K.|last2=Aboud|first2=Katherine S.|last3=Nguyen|first3=Tin Q.|last4=Cutting|first4=Laurie E.|title=Applying a network framework to the neurobiology of reading and dyslexia|journal=Journal of Neurodevelopmental Disorders|date=13 December 2018|volume=10|issue=1|page=37|doi=10.1186/s11689-018-9251-z|pmid=30541433|pmc=6291929 }}</ref><br />
<br />
<br />
*The default mode network is active when an individual is awake and at rest. It preferentially activates when individuals focus on internally-oriented tasks such as daydreaming, envisioning the future, retrieving memories, and theory of mind. It is negatively correlated with brain systems that focus on external visual signals. It is the most widely researched network.<br />
<br />
<br />
=<nowiki>= = 默认模式(Medial frontoparietal) = =</nowiki>= <br />
*默认模式网络在个人清醒和休息时处于活动状态。当个体专注于面向内部的任务时，比如做白日梦、展望未来、回忆和心理理论，它就会优先激活。它与专注于外部视觉信号的大脑系统负相关。它是研究最广泛的网络。<br />
<br />
===Salience (Midcingulo-Insular) ===<br />
{{Main|Salience network}}<br />
*The salience network consists of several structures, including the anterior (bilateral) insula, dorsal anterior cingulate cortex, and three subcortical structures which are the ventral striatum, substantia nigra/ventral tegmental region.<ref>{{Cite journal|last1=Steimke|first1=Rosa|last2=Nomi|first2=Jason S.|last3=Calhoun|first3=Vince D.|last4=Stelzel|first4=Christine|last5=Paschke|first5=Lena M.|last6=Gaschler|first6=Robert|last7=Goschke|first7=Thomas|last8=Walter|first8=Henrik|last9=Uddin|first9=Lucina Q.|date=2017-12-01|title=Salience network dynamics underlying successful resistance of temptation|journal=Social Cognitive and Affective Neuroscience|language=en|volume=12|issue=12|pages=1928–1939|doi=10.1093/scan/nsx123|pmid=29048582|pmc=5716209|issn=1749-5016|doi-access=free}}</ref><ref name=":0">{{Citation|last=Menon|first=V.|title=Brain Mapping|chapter=Salience Network|date=2015-01-01|chapter-url=http://www.sciencedirect.com/science/article/pii/B978012397025100052X|pages=597–611|editor-last=Toga|editor-first=Arthur W.|publisher=Academic Press|isbn=978-0-12-397316-0|access-date=2019-12-08|doi=10.1016/B978-0-12-397025-1.00052-X}}</ref> It plays the key role of monitoring the [[Salience (neuroscience)|salience]] of external inputs and internal brain events.<ref name="Riedl" /><ref name="Bressler" /><ref name="Bassett" /><ref name="Yuan" /><ref name="Heine" /><ref name="Yeo" /><ref name="Shafiei" /> Specifically, it aids in directing attention by identifying important biological and cognitive events.<ref name=":0" /><ref name="Bailey" /><br />
*This network includes the ventral attention network, which primarily includes the [[temporoparietal junction]] and the ventral [[frontal cortex]] of the right hemisphere.<ref name="Uddin2019" /><ref name="Vossel" /> These areas respond when behaviorally relevant stimuli occur unexpectedly.<ref name="Vossel" /> The ventral attention network is inhibited during focused attention in which top-down processing is being used, such as when visually searching for something. This response may prevent goal-driven attention from being distracted by non-relevant stimuli. It becomes active again when the target or relevant information about the target is found.<ref name="Vossel" /><ref>{{Cite journal|last1=Shulman|first1=Gordon L.|last2=McAvoy|first2=Mark P.|last3=Cowan|first3=Melanie C.|last4=Astafiev|first4=Serguei V.|last5=Tansy|first5=Aaron P.|last6=d'Avossa|first6=Giovanni|last7=Corbetta|first7=Maurizio|date=2003-11-01|title=Quantitative Analysis of Attention and Detection Signals During Visual Search|journal=Journal of Neurophysiology|volume=90|issue=5|pages=3384–3397|doi=10.1152/jn.00343.2003|pmid=12917383|issn=0022-3077}}</ref><br />
<br />
<br />
*The salience network consists of several structures, including the anterior (bilateral) insula, dorsal anterior cingulate cortex, and three subcortical structures which are the ventral striatum, substantia nigra/ventral tegmental region. It plays the key role of monitoring the salience of external inputs and internal brain events. Specifically, it aids in directing attention by identifying important biological and cognitive events.<br />
*This network includes the ventral attention network, which primarily includes the temporoparietal junction and the ventral frontal cortex of the right hemisphere. These areas respond when behaviorally relevant stimuli occur unexpectedly. The ventral attention network is inhibited during focused attention in which top-down processing is being used, such as when visually searching for something. This response may prevent goal-driven attention from being distracted by non-relevant stimuli. It becomes active again when the target or relevant information about the target is found.<br />
<br />
突显网络由前(双)岛、前扣带回皮层和腹侧纹状体、黑质/腹侧被盖区3个皮层下结构组成。它起着监测外部输入和内部脑事件的显著性的关键作用。具体来说，它通过识别重要的生物学和认知事件来帮助引导注意力。<br />
*这个网络包括腹侧注意网络，主要包括右半球的颞顶联合区和腹侧额叶皮层。当行为相关的刺激意外发生时，这些区域会做出反应。腹侧注意网络在使用自上而下加工的集中注意过程中被抑制，例如在视觉搜索某物时。这种反应可以防止目标驱动的注意力被非相关的刺激分散。当找到目标或关于目标的相关信息时，它再次激活。<br />
<br />
===Attention (Dorsal frontoparietal) ===<br />
{{Main|Dorsal attention network}}<br />
*This network is involved in the voluntary, top-down deployment of attention.<ref name="Riedl" /><ref name="Yuan" /><ref name="Bell" /><ref name="Yeo" /><ref name="Shafiei" /><ref name="Vossel">{{cite journal|last1=Vossel|first1=Simone|last2=Geng|first2=Joy J.|last3=Fink|first3=Gereon R.|title=Dorsal and Ventral Attention Systems: Distinct Neural Circuits but Collaborative Roles|journal=The Neuroscientist|date=2014|volume=20|issue=2|pages=150–159|doi=10.1177/1073858413494269|pmid=23835449|pmc=4107817}}</ref><ref name="Hutton">{{cite journal|last1=Hutton|first1=John S.|last2=Dudley|first2=Jonathan|last3=Horowitz-Kraus|first3=Tzipi|last4=DeWitt|first4=Tom|last5=Holland|first5=Scott K.|title=Functional Connectivity of Attention, Visual, and Language Networks During Audio, Illustrated, and Animated Stories in Preschool-Age Children|journal=Brain Connectivity|date=1 September 2019|volume=9|issue=7|pages=580–592|doi=10.1089/brain.2019.0679|pmid=31144523|pmc=6775495|ref=Hutton}}</ref> Within the dorsal attention network, the intraparietal sulcus and frontal eye fields influence the visual areas of the brain. These influencing factors allow for the orientation of attention.<ref>{{Cite journal|last1=Fox|first1=Michael D.|last2=Corbetta|first2=Maurizio|last3=Snyder|first3=Abraham Z.|last4=Vincent|first4=Justin L.|last5=Raichle|first5=Marcus E.|date=2006-06-27|title=Spontaneous neuronal activity distinguishes human dorsal and ventral attention systems|journal=Proceedings of the National Academy of Sciences|language=en|volume=103|issue=26|pages=10046–10051|doi=10.1073/pnas.0604187103|issn=0027-8424|pmid=16788060|pmc=1480402|bibcode=2006PNAS..10310046F|doi-access=free}}</ref><ref name="Vossel" /><ref name="Bailey" /><br />
<br />
<br />
*This network is involved in the voluntary, top-down deployment of attention. Within the dorsal attention network, the intraparietal sulcus and frontal eye fields influence the visual areas of the brain. These influencing factors allow for the orientation of attention.<br />
<br />
=<nowiki>= = 注意力(背侧额顶骨) = =</nowiki>= <br />
*这个网络参与了自发的、自上而下的注意力分配。在背侧注意网络中，顶内沟和额眼影响大脑的视觉区域。这些影响因素决定了注意力的方向。<br />
<br />
===Control (Lateral frontoparietal)===<br />
{{Main|Frontoparietal network}}<br />
*This network initiates and modulates cognitive control and comprises 18 sub-regions of the brain.<ref>{{Cite journal|last1=Scolari|first1=Miranda|last2=Seidl-Rathkopf|first2=Katharina N|last3=Kastner|first3=Sabine|date=2015-02-01|title=Functions of the human frontoparietal attention network: Evidence from neuroimaging|journal=Current Opinion in Behavioral Sciences|series=Cognitive control|volume=1|pages=32–39|doi=10.1016/j.cobeha.2014.08.003|issn=2352-1546|pmid=27398396|pmc=4936532}}</ref> There is a strong correlation between fluid intelligence and the involvement of the fronto-parietal network with other networks.<ref>{{Cite journal|last1=Marek|first1=Scott|last2=Dosenbach|first2=Nico U. F.|date=June 2018|title=The frontoparietal network: function, electrophysiology, and importance of individual precision mapping|journal=Dialogues in Clinical Neuroscience|volume=20|issue=2|pages=133–140|doi=10.31887/DCNS.2018.20.2/smarek|issn=1294-8322|pmc=6136121|pmid=30250390}}</ref><br />
<br />
<br />
*This network initiates and modulates cognitive control and comprises 18 sub-regions of the brain. There is a strong correlation between fluid intelligence and the involvement of the fronto-parietal network with other networks.<br />
<br />
=<nowiki>= = 控制(侧额顶骨) = =</nowiki>= <br />
*这个网络启动和调节认知控制，包括大脑的18个亚区。在流体智力和额顶网络与其他网络的参与之间有很强的相关性。<br />
<br />
*Versions of this network have also been called the central executive (or executive control) network and the cognitive control network.<ref name="Uddin2019" /><br />
<br />
*Versions of this network have also been called the central executive (or executive control) network and the cognitive control network.<br />
<br />
<br />
*这种网络的版本也被称为中央执行(或执行控制)网络和认知控制网络。<br />
<br />
===Sensorimotor or Somatomotor (Pericentral)===<br />
{{Main|Sensorimotor network}}<br />
*This network processes somatosensory information and coordinates motion.<ref name="Heine" /><ref name="Yeo" /><ref name="Shafiei" /><ref name="Bassett" /><ref name="Yuan" /> The [[auditory cortex]] may be included.<ref name="Uddin2019" /><ref name="Yeo" /><br />
<br />
<br />
*This network processes somatosensory information and coordinates motion. The auditory cortex may be included.<br />
<br />
===Sensorimotor or Somatomotor (Pericentral)===<br />
<br />
<br />
*This network processes somatosensory information and coordinates motion.听觉皮层可能也包括在内。<br />
<br />
===Visual (Occipital)===<br />
{{See|Visual cortex}}<br />
*This network handles visual information processing.<ref name="Yang">{{cite journal|last1=Yang|first1=Yan-li|last2=Deng|first2=Hong-xia|last3=Xing|first3=Gui-yang|last4=Xia|first4=Xiao-luan|last5=Li|first5=Hai-fang|title=Brain functional network connectivity based on a visual task: visual information processing-related brain regions are significantly activated in the task state|journal=Neural Regeneration Research|date=2015|volume=10|issue=2|pages=298–307|doi=10.4103/1673-5374.152386|pmid=25883631|pmc=4392680 }}</ref><br />
<br />
<br />
*This network handles visual information processing.<br />
<br />
= = = = = = = = = = = = = 这个网络处理视觉信息。<br />
<br />
==Other networks==<br />
Different methods and data have identified several other brain networks, many of which greatly overlap or are subsets of more well-characterized core networks.<ref name="Uddin2019" /><br />
*Limbic<ref name="Bassett" /><ref name="Yeo" /><ref name="Bailey" /><br />
*Auditory<ref name="Yuan" /><ref name="Heine" /><br />
*Right/left executive<ref name="Yuan" /><ref name="Heine" /><br />
*Cerebellar<ref name="Bell" /><ref name="Heine" /><br />
*Spatial attention<ref name="Riedl" /><ref name="Bressler" /><br />
*Language<ref name="Bressler" /><ref name="Hutton" /><br />
*Lateral visual<ref name="Yuan" /><ref name="Bell" /><ref name="Heine" /><br />
*Temporal<ref name="Yeo" /><ref name="Shafiei" /><br />
*Visual perception/imagery<ref name="Hutton" /><br />
<br />
Different methods and data have identified several other brain networks, many of which greatly overlap or are subsets of more well-characterized core networks.<br />
*Limbic<br />
*Auditory<br />
*Right/left executive<br />
*Cerebellar<br />
*Spatial attention<br />
*Language<br />
*Lateral visual<br />
*Temporal<br />
*Visual perception/imagery<br />
<br />
其他网络不同的方法和数据已经确定了其他几个大脑网络，其中许多网络极大地重叠或者是更具特色的核心网络的子集。<br />
*边缘<br />
*听觉<br />
*右/左执行<br />
* 小脑<br />
* 空间注意<br />
*语言<br />
* 外侧视觉<br />
*颞视知觉/图像<br />
<br />
==See also==<br />
*[[Complex network]]<br />
*[[Neural network]]<br />
<br />
*Complex network<br />
*Neural network<br />
<br />
=<nowiki>= = =</nowiki>= <br />
*复杂网络<br />
*<br />
*神经网络<br />
*<br />
<br />
==References==<br />
{{reflist|30em}}<br />
<br />
{{Human connectomics}}<br />
<br />
[[Category:Neuroscience]]<br />
[[Category:Neural coding]]<br />
[[Category:Neural circuits]]<br />
[[Category:Neurophysiology]]<br />
<br />
Category:Neuroscience<br />
Category:Neural coding<br />
Category:Neural circuits<br />
Category:Neurophysiology<br />
<br />
类别: 神经科学类别: 神经编码类别: 神经回路类别: 神经生理学<br />
<br />
<noinclude><br />
<br />
<small>This page was moved from [[wikipedia:en:Large-scale brain network]]. Its edit history can be viewed at [[大规模脑网络/edithistory]]</small></noinclude><br />
<br />
[[Category:待整理页面]]</div>Shenky20https://wiki.swarma.org/index.php?title=%E5%A4%A7%E8%A7%84%E6%A8%A1%E8%84%91%E7%BD%91%E7%BB%9C&diff=29972大规模脑网络2022-04-03T12:55:25Z<p>Shenky20：</p>
<hr />
<div>此词条由神经动力学模型读书会词条梳理志愿者Shenky20翻译审校，翻译字数共800，未经专家审核，带来阅读不便，请见谅。<br />
<br />
{{Short description|Collections of brain regions}}<br />
'''Large-scale brain networks''' (also known as '''intrinsic brain networks''') are collections of widespread [[brain regions]] showing [[Resting state fMRI#Functional|functional connectivity]] by statistical analysis of the [[Functional magnetic resonance imaging|fMRI]] [[BOLD signal]]<ref name=Riedl>{{cite journal|last1=Riedl|first1=Valentin|last2=Utz|first2=Lukas|last3=Castrillón|first3=Gabriel|last4=Grimmer|first4=Timo|last5=Rauschecker|first5=Josef P.|last6=Ploner|first6=Markus|last7=Friston|first7=Karl J.|last8=Drzezga|first8=Alexander|last9=Sorg|first9=Christian|title=Metabolic connectivity mapping reveals effective connectivity in the resting human brain|journal=PNAS|date=January 12, 2016|volume=113|issue=2|pages=428–433|doi=10.1073/pnas.1513752113|pmid=26712010|pmc=4720331|bibcode=2016PNAS..113..428R|doi-access=free}}</ref> or other recording methods such as [[Electroencephalography|EEG]],<ref>{{Cite journal|last1=Foster|first1=Brett L.|last2=Parvizi|first2=Josef|date=2012-03-01|title=Resting oscillations and cross-frequency coupling in the human posteromedial cortex|journal=NeuroImage|volume=60|issue=1|pages=384–391|doi=10.1016/j.neuroimage.2011.12.019|pmid=22227048|issn=1053-8119|pmc=3596417}}</ref> [[Positron emission tomography|PET]]<ref>{{Cite journal|last1=Buckner|first1=Randy L.|last2=Andrews‐Hanna|first2=Jessica R.|last3=Schacter|first3=Daniel L.|date=2008|title=The Brain's Default Network|journal=Annals of the New York Academy of Sciences|language=en|volume=1124|issue=1|pages=1–38|doi=10.1196/annals.1440.011|pmid=18400922|issn=1749-6632|bibcode=2008NYASA1124....1B|s2cid=3167595}}</ref> and [[Magnetoencephalography|MEG]].<ref>{{Cite journal|last1=Morris|first1=Peter G.|last2=Smith|first2=Stephen M.|last3=Barnes|first3=Gareth R.|last4=Stephenson|first4=Mary C.|last5=Hale|first5=Joanne R.|last6=Price|first6=Darren|last7=Luckhoo|first7=Henry|last8=Woolrich|first8=Mark|last9=Brookes|first9=Matthew J.|date=2011-10-04|title=Investigating the electrophysiological basis of resting state networks using magnetoencephalography|journal=Proceedings of the National Academy of Sciences|language=en|volume=108|issue=40|pages=16783–16788|doi=10.1073/pnas.1112685108|issn=0027-8424|pmid=21930901|pmc=3189080|bibcode=2011PNAS..10816783B|doi-access=free}}</ref> An emerging paradigm in neuroscience is that cognitive tasks are performed not by individual brain regions working in isolation but by networks consisting of several discrete brain regions that are said to be "functionally connected". Functional connectivity networks may be found using algorithms such as [[cluster analysis]], spatial [[independent component analysis]] (ICA), seed based, and others.<ref name="Petersen">{{cite journal|last1=Petersen|first1=Steven|last2=Sporns|first2=Olaf|title=Brain Networks and Cognitive Architectures|journal=Neuron|date=October 2015|volume=88|issue=1|pages=207–219|doi=10.1016/j.neuron.2015.09.027|pmid=26447582|pmc=4598639 }}</ref> Synchronized brain regions may also be identified using long-range synchronization of the EEG, MEG, or other dynamic brain signals.<ref name=Bressler>{{cite journal|last1=Bressler|first1=Steven L.|last2=Menon|first2=Vinod|s2cid=5967761|title=Large scale brain networks in cognition: emerging methods and principles|journal=Trends in Cognitive Sciences|date=June 2010|volume=14|issue=6|pages=233–290|doi=10.1016/j.tics.2010.04.004|url=http://www.cell.com/trends/cognitive-sciences/issue?pii=S1364-6613(10)X0005-5|accessdate=24 January 2016|pmid=20493761}}</ref><br />
<br />
<br />
Large-scale brain networks (also known as intrinsic brain networks) are collections of widespread brain regions showing functional connectivity by statistical analysis of the fMRI BOLD signal or other recording methods such as EEG, PET and MEG. An emerging paradigm in neuroscience is that cognitive tasks are performed not by individual brain regions working in isolation but by networks consisting of several discrete brain regions that are said to be "functionally connected". Functional connectivity networks may be found using algorithms such as cluster analysis, spatial independent component analysis (ICA), seed based, and others. Synchronized brain regions may also be identified using long-range synchronization of the EEG, MEG, or other dynamic brain signals.<br />
<br />
大规模的大脑网络(也称为内在大脑网络)是通过功能性磁共振成像 BOLD 信号的统计分析或其他记录方法(如 EEG、 PET 和 MEG)显示功能连接的广泛的大脑区域的集合。神经科学中一个新出现的范式是，认知任务不是由独立工作的单个大脑区域执行的，而是由几个被称为“功能连接”的离散大脑区域组成的网络执行的。功能连接网络可以通过算法来发现，比如数据聚类、空间独立元素分析、种子等等。同步的大脑区域也可以用脑电图、脑磁图或其他动态大脑信号的远程同步来识别。<br />
<br />
The set of identified brain areas that are linked together in a large-scale network varies with cognitive function.<ref name="Bressler2">{{cite journal|last1=Bressler|first1=Steven L.|title=Neurocognitive networks|journal=Scholarpedia|volume=3|issue=2|pages=1567|doi=10.4249/scholarpedia.1567|year=2008|bibcode=2008SchpJ...3.1567B|doi-access=free}}</ref> When the cognitive state is not explicit (i.e., the subject is at "rest"), the large-scale brain network is a [[Resting state fMRI|resting state]] network (RSN). As a physical system with graph-like properties,<ref name="Bressler" /> a large-scale brain network has both nodes and edges and cannot be identified simply by the co-activation of brain areas. In recent decades, the analysis of brain networks was made feasible by advances in imaging techniques as well as new tools from [[graph theory]] and [[Dynamical systems theory|dynamical systems]].<br />
<br />
The set of identified brain areas that are linked together in a large-scale network varies with cognitive function. When the cognitive state is not explicit (i.e., the subject is at "rest"), the large-scale brain network is a resting state network (RSN). As a physical system with graph-like properties, a large-scale brain network has both nodes and edges and cannot be identified simply by the co-activation of brain areas. In recent decades, the analysis of brain networks was made feasible by advances in imaging techniques as well as new tools from graph theory and dynamical systems.<br />
<br />
在一个大规模的网络中，连接在一起的大脑区域的集合因认知功能的不同而不同。当认知状态不明确(即主体处于“静止”状态)时，大规模的脑网络是一个静息状态网络(RSN)。作为一个具有图形特征的物理系统，大规模的脑网络既有节点又有边，不能简单地通过脑区的共同激活来识别。近几十年来，由于成像技术的进步以及图论和动力学系统的新工具，脑网络的分析变得可行。<br />
<br />
Large-scale brain networks are identified by their function and provide a coherent framework for understanding [[cognition]] by offering a neural model of how different cognitive functions emerge when different sets of brain regions join together as self-organized coalitions. The number and composition of the coalitions will vary with the algorithm and parameters used to identify them.<ref name="Yeo" /><ref>{{cite journal|last1=Abou Elseoud|first1=Ahmed|last2=Littow|first2=Harri|last3=Remes|first3=Jukka|last4=Starck|first4=Tuomo|last5=Nikkinen|first5=Juha|last6=Nissilä|first6=Juuso|last7=Timonen|first7=Markku|last8=Tervonen|first8=Osmo|last9=Kiviniemi1|first9=Vesa|title=Group-ICA Model Order Highlights Patterns of Functional Brain Connectivity|journal= Frontiers in Systems Neuroscience|date=2011-06-03|volume=5|pages=37|doi=10.3389/fnsys.2011.00037|pmid=21687724|pmc=3109774 |doi-access=free}}</ref> In one model, there is only the [[default mode network]] and the [[task-positive network]], but most current analyses show several networks, from a small handful to 17.<ref name="Yeo" /> The most common and stable networks are enumerated below. The regions participating in a functional network may be dynamically reconfigured.<ref name="Petersen" /><ref name="Bassett">{{cite journal|last1=Bassett|first1=Daniella|last2=Bertolero|first2=Max|title=How Matter Becomes Mind|journal=Scientific American|date=July 2019|volume=321|issue=1|page=32|url=https://www.scientificamerican.com/|accessdate=23 June 2019 }}</ref><br />
<br />
Large-scale brain networks are identified by their function and provide a coherent framework for understanding cognition by offering a neural model of how different cognitive functions emerge when different sets of brain regions join together as self-organized coalitions. The number and composition of the coalitions will vary with the algorithm and parameters used to identify them. In one model, there is only the default mode network and the task-positive network, but most current analyses show several networks, from a small handful to 17. The most common and stable networks are enumerated below. The regions participating in a functional network may be dynamically reconfigured.<br />
<br />
大规模的大脑网络是通过其功能识别的，并通过提供一个神经模型，说明当不同的大脑区域组合在一起形成自组织的联盟时，不同的认知功能是如何产生的，从而为理解认知提供了一个连贯的框架。联盟的数量和组成将根据识别联盟的算法和参数而有所不同。在一个模型中，只有默认模式网络和任务正向网络，但目前大多数分析显示，从少数到17个网络。下面列举了最常见和最稳定的网络。可以动态地重新配置参与功能网络的区域。<br />
<br />
Disruptions in activity in various networks have been implicated in neuropsychiatric disorders such as [[Depression (mood)|depression]], [[Alzheimer's disease|Alzheimer's]], [[Autism-spectrum disorder|autism spectrum disorder]], [[schizophrenia]], [[ADHD]]<ref>{{cite journal |last1=Griffiths |first1=Kristi R. |last2=Braund |first2=Taylor A. |last3=Kohn |first3=Michael R. |last4=Clarke |first4=Simon |last5=Williams |first5=Leanne M. |last6=Korgaonkar |first6=Mayuresh S. |title=Structural brain network topology underpinning ADHD and response to methylphenidate treatment |journal=Translational Psychiatry |date=2 March 2021 |volume=11 |issue=1 |pages=1–9 |doi=10.1038/s41398-021-01278-x | pmc=7925571 |pmid=33654073 |url=https://www.nature.com/articles/s41398-021-01278-x#citeas |access-date=16 November 2021}}</ref> and [[bipolar disorder]].<ref>{{Cite journal|url=https://www.researchgate.net/publication/51639686|title=Large-scale brain networks and psychopathology: A unifying triple network model|last=Menon|first=Vinod|s2cid=26653572|journal=Trends in Cognitive Sciences|date=2011-09-09|volume=15|issue=10|pages=483–506|doi=10.1016/j.tics.2011.08.003|pmid=21908230}}</ref><br />
<br />
Disruptions in activity in various networks have been implicated in neuropsychiatric disorders such as depression, Alzheimer's, autism spectrum disorder, schizophrenia, ADHD and bipolar disorder.<br />
<br />
各种网络活动的中断牵连到神经精神疾病，如抑郁症、老年痴呆症、自闭症光谱、精神分裂症、多动症和躁郁症。<br />
<br />
== Core networks ==<br />
[[File:Heine2012x3010.png|thumb|An example that identified 10 large-scale brain networks from [[resting state fMRI]] activity through [[independent component analysis]].<ref name="Heine" />|链接=Special:FilePath/Heine2012x3010.png]]<br />
Because brain networks can be identified at various different resolutions and with various different neurobiological properties, there is no such thing as a universal atlas of brain networks that fits all circumstances.<ref>{{cite journal|last1=Eickhoff|first1=SB|last2=Yeo|first2=BTT|last3=Genon|first3=S|title=Imaging-based parcellations of the human brain.|journal=Nature Reviews. Neuroscience|date=November 2018|volume=19|issue=11|pages=672–686|doi=10.1038/s41583-018-0071-7|pmid=30305712|s2cid=52954265|url=http://juser.fz-juelich.de/record/856633/files/Eickhoff_Yeo_Genon_NRN_MainManuscriptInclFigures.pdf}}</ref> While acknowledging this problem, Uddin, Yeo, and Spreng proposed in 2019<ref name="Uddin2019">{{cite journal|last1=Uddin|first1=LQ|last2=Yeo|first2=BTT|last3=Spreng|first3=RN|title=Towards a Universal Taxonomy of Macro-scale Functional Human Brain Networks.|journal=Brain Topography|date=November 2019|volume=32|issue=6|pages=926–942|doi=10.1007/s10548-019-00744-6|pmid=31707621|pmc=7325607}}</ref> that the following six networks should be defined as core networks based on converging evidences from multiple studies<ref>{{cite journal|last1=Doucet|first1=GE|last2=Lee|first2=WH|last3=Frangou|first3=S|title=Evaluation of the spatial variability in the major resting-state networks across human brain functional atlases.|journal=Human Brain Mapping|date=2019-10-15|volume=40|issue=15|pages=4577–4587|doi=10.1002/hbm.24722|pmid=31322303|pmc=6771873}}</ref><ref name="Yeo" /><ref>{{cite journal|last1=Smith|first1=SM|last2=Fox|first2=PT|last3=Miller|first3=KL|last4=Glahn|first4=DC|last5=Fox|first5=PM|last6=Mackay|first6=CE|last7=Filippini|first7=N|last8=Watkins|first8=KE|last9=Toro|first9=R|last10=Laird|first10=AR|last11=Beckmann|first11=CF|title=Correspondence of the brain's functional architecture during activation and rest.|journal=Proceedings of the National Academy of Sciences of the United States of America|date=2009-08-04|volume=106|issue=31|pages=13040–5|doi=10.1073/pnas.0905267106|pmid=19620724|pmc=2722273|bibcode=2009PNAS..10613040S|doi-access=free}}</ref> to facilitate communication between researchers.<br />
<br />
<br />
Because brain networks can be identified at various different resolutions and with various different neurobiological properties, there is no such thing as a universal atlas of brain networks that fits all circumstances. While acknowledging this problem, Uddin, Yeo, and Spreng proposed in 2019 that the following six networks should be defined as core networks based on converging evidences from multiple studies to facilitate communication between researchers.<br />
<br />
因为大脑网络可以用不同的分辨率和不同的神经生物学特性来识别，所以没有适合所有情况的通用大脑网络图谱。在承认这个问题的同时，Uddin，Yeo，和 Spreng 在2019年提出，以下六个网络应该被定义为核心网络，基于来自多个研究的聚合证据，以促进研究人员之间的交流。<br />
<br />
=== Default Mode (Medial frontoparietal) ===<br />
{{Main|Default mode network}}<br />
*The default mode network is active when an individual is awake and at rest. It preferentially activates when individuals focus on internally-oriented tasks such as daydreaming, envisioning the future, retrieving memories, and [[theory of mind]]. It is negatively correlated with brain systems that focus on external visual signals. It is the most widely researched network.<ref name="Bressler" /><ref name="Bassett" /><ref>{{Cite journal|date=2012-08-15|title=The serendipitous discovery of the brain's default network|journal=NeuroImage|language=en|volume=62|issue=2|pages=1137–1145|doi=10.1016/j.neuroimage.2011.10.035|pmid=22037421|issn=1053-8119|last1=Buckner|first1=Randy L.|s2cid=9880586}}</ref><ref name="Riedl" /><ref name="Yuan"><br />
{{cite journal|last1=Yuan|first1=Rui|last2=Di|first2=Xin|last3=Taylor|first3=Paul A.|last4=Gohel|first4=Suril|last5=Tsai|first5=Yuan-Hsiung|last6=Biswal|first6=Bharat B.|title=Functional topography of the thalamocortical system in human|journal=Brain Structure and Function|date=30 April 2015|doi=10.1007/s00429-015-1018-7|pmid=25924563|pmc=6363530|volume=221|issue=4|pages=1971–1984}}</ref><ref name="Bell">{{cite journal|last1=Bell|first1=Peter T.|last2=Shine|first2=James M.|title=Estimating Large-Scale Network Convergence in the Human Functional Connectome|journal=Brain Connectivity|date=2015-11-09|volume=5|issue=9|doi=10.1089/brain.2015.0348|pmid=26005099|pages=565–74}}</ref><ref name="Heine">{{cite journal|last1=Heine|first1=Lizette|last2=Soddu|first2=Andrea|last3=Gomez|first3=Francisco|last4=Vanhaudenhuyse|first4=Audrey|last5=Tshibanda|first5=Luaba|last6=Thonnard|first6=Marie|last7=Charland-Verville|first7=Vanessa|last8=Kirsch|first8=Murielle|last9=Laureys|first9=Steven|last10=Demertzi|first10=Athena|title=Resting state networks and consciousness. Alterations of multiple resting state network connectivity in physiological, pharmacological and pathological consciousness states.|journal=Frontiers in Psychology|date=2012|volume=3|pages=295|doi=10.3389/fpsyg.2012.00295|pmid=22969735|pmc=3427917|doi-access=free}}</ref><ref name="Yeo">{{cite journal|last1=Yeo|first1=B. T. Thomas|last2=Krienen|first2=Fenna M.|last3=Sepulcre|first3=Jorge|last4=Sabuncu|first4=Mert R.|last5=Lashkari|first5=Danial|last6=Hollinshead|first6=Marisa|last7=Roffman|first7=Joshua L.|last8=Smoller|first8=Jordan W.|last9=Zöllei|first9=Lilla|last10=Polimeni|first10=Jonathan R.|last11=Fischl|first11=Bruce|last12=Liu|first12=Hesheng|last13=Buckner|first13=Randy L.|title=The organization of the human cerebral cortex estimated by intrinsic functional connectivity|journal=Journal of Neurophysiology|date=2011-09-01|volume=106|issue=3|pages=1125–1165|doi=10.1152/jn.00338.2011|pmid=21653723|pmc=3174820|bibcode=2011NatSD...2E0031H }}</ref><ref name="Shafiei">{{cite journal|last1=Shafiei|first1=Golia|last2=Zeighami|first2=Yashar|last3=Clark|first3=Crystal A.|last4=Coull|first4=Jennifer T.|last5=Nagano-Saito|first5=Atsuko|last6=Leyton|first6=Marco|last7=Dagher|first7=Alain|last8=Mišić|first8=Bratislav|title=Dopamine Signaling Modulates the Stability and Integration of Intrinsic Brain Networks|journal=Cerebral Cortex|date=2018-10-01|volume=29|issue=1|pages=397–409|doi=10.1093/cercor/bhy264|pmid=30357316|pmc=6294404 }}</ref><ref name="Bailey">{{cite journal|last1=Bailey|first1=Stephen K.|last2=Aboud|first2=Katherine S.|last3=Nguyen|first3=Tin Q.|last4=Cutting|first4=Laurie E.|title=Applying a network framework to the neurobiology of reading and dyslexia|journal=Journal of Neurodevelopmental Disorders|date=13 December 2018|volume=10|issue=1|page=37|doi=10.1186/s11689-018-9251-z|pmid=30541433|pmc=6291929 }}</ref><br />
<br />
<br />
*The default mode network is active when an individual is awake and at rest. It preferentially activates when individuals focus on internally-oriented tasks such as daydreaming, envisioning the future, retrieving memories, and theory of mind. It is negatively correlated with brain systems that focus on external visual signals. It is the most widely researched network.<br />
<br />
<br />
= = = 默认模式(Medial frontoparietal) = = = <br />
* 默认模式网络在个人清醒和休息时处于活动状态。当个体专注于面向内部的任务时，比如做白日梦、展望未来、回忆和心理理论，它就会优先激活。它与专注于外部视觉信号的大脑系统负相关。它是研究最广泛的网络。<br />
<br />
=== Salience (Midcingulo-Insular) ===<br />
{{Main|Salience network}}<br />
*The salience network consists of several structures, including the anterior (bilateral) insula, dorsal anterior cingulate cortex, and three subcortical structures which are the ventral striatum, substantia nigra/ventral tegmental region.<ref>{{Cite journal|last1=Steimke|first1=Rosa|last2=Nomi|first2=Jason S.|last3=Calhoun|first3=Vince D.|last4=Stelzel|first4=Christine|last5=Paschke|first5=Lena M.|last6=Gaschler|first6=Robert|last7=Goschke|first7=Thomas|last8=Walter|first8=Henrik|last9=Uddin|first9=Lucina Q.|date=2017-12-01|title=Salience network dynamics underlying successful resistance of temptation|journal=Social Cognitive and Affective Neuroscience|language=en|volume=12|issue=12|pages=1928–1939|doi=10.1093/scan/nsx123|pmid=29048582|pmc=5716209|issn=1749-5016|doi-access=free}}</ref><ref name=":0">{{Citation|last=Menon|first=V.|title=Brain Mapping|chapter=Salience Network|date=2015-01-01|chapter-url=http://www.sciencedirect.com/science/article/pii/B978012397025100052X|pages=597–611|editor-last=Toga|editor-first=Arthur W.|publisher=Academic Press|isbn=978-0-12-397316-0|access-date=2019-12-08|doi=10.1016/B978-0-12-397025-1.00052-X}}</ref> It plays the key role of monitoring the [[Salience (neuroscience)|salience]] of external inputs and internal brain events.<ref name="Riedl" /><ref name="Bressler" /><ref name="Bassett" /><ref name="Yuan" /><ref name="Heine" /><ref name="Yeo" /><ref name="Shafiei" /> Specifically, it aids in directing attention by identifying important biological and cognitive events.<ref name=":0" /><ref name="Bailey" /><br />
*This network includes the ventral attention network, which primarily includes the [[temporoparietal junction]] and the ventral [[frontal cortex]] of the right hemisphere.<ref name="Uddin2019" /><ref name="Vossel" /> These areas respond when behaviorally relevant stimuli occur unexpectedly.<ref name="Vossel" /> The ventral attention network is inhibited during focused attention in which top-down processing is being used, such as when visually searching for something. This response may prevent goal-driven attention from being distracted by non-relevant stimuli. It becomes active again when the target or relevant information about the target is found.<ref name="Vossel" /><ref>{{Cite journal|last1=Shulman|first1=Gordon L.|last2=McAvoy|first2=Mark P.|last3=Cowan|first3=Melanie C.|last4=Astafiev|first4=Serguei V.|last5=Tansy|first5=Aaron P.|last6=d'Avossa|first6=Giovanni|last7=Corbetta|first7=Maurizio|date=2003-11-01|title=Quantitative Analysis of Attention and Detection Signals During Visual Search|journal=Journal of Neurophysiology|volume=90|issue=5|pages=3384–3397|doi=10.1152/jn.00343.2003|pmid=12917383|issn=0022-3077}}</ref><br />
<br />
<br />
*The salience network consists of several structures, including the anterior (bilateral) insula, dorsal anterior cingulate cortex, and three subcortical structures which are the ventral striatum, substantia nigra/ventral tegmental region. It plays the key role of monitoring the salience of external inputs and internal brain events. Specifically, it aids in directing attention by identifying important biological and cognitive events.<br />
*This network includes the ventral attention network, which primarily includes the temporoparietal junction and the ventral frontal cortex of the right hemisphere. These areas respond when behaviorally relevant stimuli occur unexpectedly. The ventral attention network is inhibited during focused attention in which top-down processing is being used, such as when visually searching for something. This response may prevent goal-driven attention from being distracted by non-relevant stimuli. It becomes active again when the target or relevant information about the target is found.<br />
<br />
突显网络由前(双)岛、前扣带回皮层和腹侧纹状体、黑质/腹侧被盖区3个皮层下结构组成。它起着监测外部输入和内部脑事件的显著性的关键作用。具体来说，它通过识别重要的生物学和认知事件来帮助引导注意力。<br />
* 这个网络包括腹侧注意网络，主要包括右半球的颞顶联合区和腹侧额叶皮层。当行为相关的刺激意外发生时，这些区域会做出反应。腹侧注意网络在使用自上而下加工的集中注意过程中被抑制，例如在视觉搜索某物时。这种反应可以防止目标驱动的注意力被非相关的刺激分散。当找到目标或关于目标的相关信息时，它再次激活。<br />
<br />
=== Attention (Dorsal frontoparietal) ===<br />
{{Main|Dorsal attention network}}<br />
*This network is involved in the voluntary, top-down deployment of attention.<ref name="Riedl" /><ref name="Yuan" /><ref name="Bell" /><ref name="Yeo" /><ref name="Shafiei" /><ref name="Vossel">{{cite journal|last1=Vossel|first1=Simone|last2=Geng|first2=Joy J.|last3=Fink|first3=Gereon R.|title=Dorsal and Ventral Attention Systems: Distinct Neural Circuits but Collaborative Roles|journal=The Neuroscientist|date=2014|volume=20|issue=2|pages=150–159|doi=10.1177/1073858413494269|pmid=23835449|pmc=4107817}}</ref><ref name="Hutton">{{cite journal|last1=Hutton|first1=John S.|last2=Dudley|first2=Jonathan|last3=Horowitz-Kraus|first3=Tzipi|last4=DeWitt|first4=Tom|last5=Holland|first5=Scott K.|title=Functional Connectivity of Attention, Visual, and Language Networks During Audio, Illustrated, and Animated Stories in Preschool-Age Children|journal=Brain Connectivity|date=1 September 2019|volume=9|issue=7|pages=580–592|doi=10.1089/brain.2019.0679|pmid=31144523|pmc=6775495|ref=Hutton}}</ref> Within the dorsal attention network, the intraparietal sulcus and frontal eye fields influence the visual areas of the brain. These influencing factors allow for the orientation of attention.<ref>{{Cite journal|last1=Fox|first1=Michael D.|last2=Corbetta|first2=Maurizio|last3=Snyder|first3=Abraham Z.|last4=Vincent|first4=Justin L.|last5=Raichle|first5=Marcus E.|date=2006-06-27|title=Spontaneous neuronal activity distinguishes human dorsal and ventral attention systems|journal=Proceedings of the National Academy of Sciences|language=en|volume=103|issue=26|pages=10046–10051|doi=10.1073/pnas.0604187103|issn=0027-8424|pmid=16788060|pmc=1480402|bibcode=2006PNAS..10310046F|doi-access=free}}</ref><ref name="Vossel" /><ref name="Bailey" /><br />
<br />
<br />
*This network is involved in the voluntary, top-down deployment of attention. Within the dorsal attention network, the intraparietal sulcus and frontal eye fields influence the visual areas of the brain. These influencing factors allow for the orientation of attention.<br />
<br />
= = = 注意力(背侧额顶骨) = = = <br />
* 这个网络参与了自发的、自上而下的注意力分配。在背侧注意网络中，顶内沟和额眼影响大脑的视觉区域。这些影响因素决定了注意力的方向。<br />
<br />
=== Control (Lateral frontoparietal) ===<br />
{{Main|Frontoparietal network}}<br />
*This network initiates and modulates cognitive control and comprises 18 sub-regions of the brain.<ref>{{Cite journal|last1=Scolari|first1=Miranda|last2=Seidl-Rathkopf|first2=Katharina N|last3=Kastner|first3=Sabine|date=2015-02-01|title=Functions of the human frontoparietal attention network: Evidence from neuroimaging|journal=Current Opinion in Behavioral Sciences|series=Cognitive control|volume=1|pages=32–39|doi=10.1016/j.cobeha.2014.08.003|issn=2352-1546|pmid=27398396|pmc=4936532}}</ref> There is a strong correlation between fluid intelligence and the involvement of the fronto-parietal network with other networks.<ref>{{Cite journal|last1=Marek|first1=Scott|last2=Dosenbach|first2=Nico U. F.|date=June 2018|title=The frontoparietal network: function, electrophysiology, and importance of individual precision mapping|journal=Dialogues in Clinical Neuroscience|volume=20|issue=2|pages=133–140|doi=10.31887/DCNS.2018.20.2/smarek|issn=1294-8322|pmc=6136121|pmid=30250390}}</ref><br />
<br />
<br />
*This network initiates and modulates cognitive control and comprises 18 sub-regions of the brain. There is a strong correlation between fluid intelligence and the involvement of the fronto-parietal network with other networks.<br />
<br />
= = = 控制(侧额顶骨) = = = <br />
* 这个网络启动和调节认知控制，包括大脑的18个亚区。在流体智力和额顶网络与其他网络的参与之间有很强的相关性。<br />
<br />
*Versions of this network have also been called the central executive (or executive control) network and the cognitive control network.<ref name="Uddin2019" /><br />
<br />
*Versions of this network have also been called the central executive (or executive control) network and the cognitive control network.<br />
<br />
<br />
* 这种网络的版本也被称为中央执行(或执行控制)网络和认知控制网络。<br />
<br />
=== Sensorimotor or Somatomotor (Pericentral) ===<br />
{{Main|Sensorimotor network}}<br />
*This network processes somatosensory information and coordinates motion.<ref name=Heine /><ref name=Yeo /><ref name =Shafiei /><ref name=Bassett /><ref name=Yuan /> The [[auditory cortex]] may be included.<ref name="Uddin2019" /><ref name=Yeo /><br />
<br />
<br />
*This network processes somatosensory information and coordinates motion. The auditory cortex may be included.<br />
<br />
=== Sensorimotor or Somatomotor (Pericentral) ===<br />
<br />
<br />
*This network processes somatosensory information and coordinates motion.听觉皮层可能也包括在内。<br />
<br />
=== Visual (Occipital) ===<br />
{{See|Visual cortex}}<br />
*This network handles visual information processing.<ref name="Yang">{{cite journal|last1=Yang|first1=Yan-li|last2=Deng|first2=Hong-xia|last3=Xing|first3=Gui-yang|last4=Xia|first4=Xiao-luan|last5=Li|first5=Hai-fang|title=Brain functional network connectivity based on a visual task: visual information processing-related brain regions are significantly activated in the task state|journal=Neural Regeneration Research|date=2015|volume=10|issue=2|pages=298–307|doi=10.4103/1673-5374.152386|pmid=25883631|pmc=4392680 }}</ref><br />
<br />
<br />
*This network handles visual information processing.<br />
<br />
= = = = = = = = = = = = = 这个网络处理视觉信息。<br />
<br />
== Other networks ==<br />
Different methods and data have identified several other brain networks, many of which greatly overlap or are subsets of more well-characterized core networks.<ref name="Uddin2019" /><br />
* Limbic<ref name=Bassett /><ref name=Yeo /><ref name="Bailey" /><br />
* Auditory<ref name=Yuan /><ref name=Heine /><br />
* Right/left executive<ref name=Yuan /><ref name=Heine /><br />
* Cerebellar<ref name=Bell /><ref name=Heine /><br />
* Spatial attention<ref name= Riedl /><ref name=Bressler /><br />
* Language<ref name=Bressler /><ref name =Hutton /><br />
* Lateral visual<ref name="Yuan" /><ref name="Bell" /><ref name="Heine" /><br />
* Temporal<ref name=Yeo /><ref name =Shafiei /><br />
* Visual perception/imagery<ref name =Hutton /><br />
<br />
Different methods and data have identified several other brain networks, many of which greatly overlap or are subsets of more well-characterized core networks.<br />
* Limbic<br />
* Auditory<br />
* Right/left executive<br />
* Cerebellar<br />
* Spatial attention<br />
* Language<br />
* Lateral visual<br />
* Temporal<br />
* Visual perception/imagery<br />
<br />
其他网络不同的方法和数据已经确定了其他几个大脑网络，其中许多网络极大地重叠或者是更具特色的核心网络的子集。<br />
* 边缘 <br />
* 听觉 <br />
* 右/左执行 <br />
* 小脑 <br />
* 空间注意 <br />
* 语言 <br />
* 外侧视觉 <br />
* 颞视知觉/图像<br />
<br />
==See also==<br />
*[[Complex network]]<br />
*[[Neural network]]<br />
<br />
*Complex network<br />
*Neural network<br />
<br />
= = = = = <br />
* 复杂网络 <br />
* <br />
* 神经网络 <br />
* <br />
<br />
==References==<br />
{{reflist|30em}}<br />
<br />
{{Human connectomics}}<br />
<br />
[[Category:Neuroscience]]<br />
[[Category:Neural coding]]<br />
[[Category:Neural circuits]]<br />
[[Category:Neurophysiology]]<br />
<br />
Category:Neuroscience<br />
Category:Neural coding<br />
Category:Neural circuits<br />
Category:Neurophysiology<br />
<br />
类别: 神经科学类别: 神经编码类别: 神经回路类别: 神经生理学<br />
<br />
<noinclude><br />
<br />
<small>This page was moved from [[wikipedia:en:Large-scale brain network]]. Its edit history can be viewed at [[大规模脑网络/edithistory]]</small></noinclude><br />
<br />
[[Category:待整理页面]]</div>Shenky20https://wiki.swarma.org/index.php?title=%E7%94%A8%E6%88%B7:Shenky20&diff=29722用户:Shenky202022-03-27T13:26:39Z<p>Shenky20：</p>
<hr />
<div>清华大学未央书院2020级本科生<br />
<br />
请多多指教！</div>Shenky20https://wiki.swarma.org/index.php?title=%E7%94%A8%E6%88%B7:Shenky20&diff=29600用户:Shenky202022-03-25T10:59:46Z<p>Shenky20：建立内容为“清华大学未央书院2020级本科生”的新页面</p>
<hr />
<div>清华大学未央书院2020级本科生</div>Shenky20