Article(id=1329420410619277551, tenantId=1146029695717560320, journalId=1324759914146529478, issueId=1329420206000157303, articleNumber=null, orderNo=null, doi=10.7544/issn1000-1239.202550722, pmid=null, cstr=null, oa=null, hot=null, price=null, onlineType=0, articleFormat=0, articleType=null, articleTypeStr=null, receivedDate=1759161600000, receivedDateStr=2025-09-30, revisedDate=1774800000000, revisedDateStr=2026-03-30, acceptedDate=null, acceptedDateStr=null, onlineDate=1794888125328, onlineDateStr=2026-11-17, pubDate=1785513600000, pubDateStr=2026-08-01, doiRegisterDate=null, doiRegisterDateStr=null, onlineIssueDate=1794888125328, onlineIssueDateStr=2026-11-17, onlineJustAcceptDate=null, onlineJustAcceptDateStr=null, onlineFirstDate=null, onlineFirstDateStr=null, sourceXml=null, magXml=null, createTime=1794888125328, creator=13701087609, updateTime=1794888125328, updator=13701087609, issue=Issue{id=1329420206000157303, tenantId=1146029695717560320, journalId=1324759914146529478, year='2026', volume='63', issue='8', pageStart='1905', pageEnd='2181', issueExtLink='null', onlineDate='null', pubDate='1785513600000', pubDateStr='2026-08-01', beforeIssueId=null, nextIssueId=null, price=null, status=1, issueComplete=1, articleOrder=1, issueType=-1, specialIssue=null, createTime=1794888076542, creator='13701087609', updateTime=1794888076542, updator='13701087609', preIssue=null, nextIssue=null, articleTotal=null, ext=null, issueFiles=null, downloadFileDto=null}, startPage=2143, endPage=2153, ext={EN=ArticleExt(id=1329420410824798448, articleId=1329420410619277551, tenantId=1146029695717560320, journalId=1324759914146529478, language=EN, title=SpikingReLU: A High-Performance Spiking Neural Network Based on Hybrid Model Conversion, columnId=1325489891789910943, journalTitle=Journal of Computer Research and Development, columnName=Artificial Intelligence, runingTitle=null, highlight=null, articleAbstract=

Brain-inspired spiking neural networks (SNNs) have shown considerable promise due to their strong spatiotemporal encoding capabilities and event-driven computation. Current mainstream SNN approaches can be divided into two categories: converting pre-trained artificial neural networks (ANNs) into SNNs (ANN conversion methods) and directly training SNNs. However, ANN conversion methods require large time steps to reduce conversion errors, while directly trained SNNs suffer from poor representational capacity. To address these challenges, we propose a high-performance spiking neural network based on hybrid model conversion. The proposed method decouples model training and inference. During the training phase, a portion of the spiking neurons in the SNNs are strategically replaced with ReLU activation functions to enhance the feature learning ability, thereby constructing a high-performance hybrid model that integrates both ANN and SNN components. Since the ANN part of the hybrid model cannot perform event-driven computation, we further introduce a reparameterization technique to convert the entire inference process of the hybrid model into event-driven computation without performance loss. Therefore, the proposed method combines the advantages of high performance and event-driven computation. Experimental results show that the proposed method achieves classification accuracies of 97.31% and 83.34% on CIFAR-10 and CIFAR-100, respectively, 70.89% on ImageNet, and 82.71% on the neuromorphic dataset CIFAR10-DVS, outperforming state-of-the-art methods.

, authors=Yufei Wang1, Jingyue Zhao2, Ge Zhang3, Bin Fan4, Zhijie Yang2, Zhong Wan2, Yuchao Dai4, Lei Wang2, *, authorsList=Yufei Wang, Jingyue Zhao, Ge Zhang, Bin Fan, Zhijie Yang, Zhong Wan, Yuchao Dai, Lei Wang, authorCompany=null, correspAuthors=Lei Wang, authorNote=

born in 1995. PhD. His main research interests include machine vision and neuromorphic computing

born in 1994. PhD, assistant professor. Member of CCF. Her main research interests include neuromorphic computing, brain-inspired robotics, and embodied intelligence

born in 1996. PhD, assistant professor. Her main research interests include hyperspectral image interpretation and multimodal image fusion

born in 1994. PhD, professor, PhD supervisor. His main research interests include machine vision, computational photography, and 3D reconstruction

born in 1995. PhD. His main research interests include neuromorphic computing and computer architecture

born in 1989. PhD, associate professor. Her main research interest includes neuromorphic computing

born in 1982. PhD, professor, PhD supervisor. Senior member of CCF. His main research interests include machine vision and embodied intelligence

born in 1977. PhD, professor, PhD supervisor. Distinguished member of CCF. Her main research interests include computer architecture and intelligent chips

, correspAuthorsNote=null, copyrightStatement=Copyright © 2026 Journal of Computer Research and Development. All rights reserved., copyrightOwner=null, extLink=null, articleAbsUrl=null, sourceXml=null, magXml=null, pdfUrl=null, pdf=null, pdfFileSize=null, pdfExtLink=null, richHtmlUrl=null, mobilePdfUrl=null, reviewReport=null, pdfFirstPage=null, abstractGraph=null, abstractGraphContent=null, abstractVideo=null, citation=null, cebUrl=null, magXmlContent=null, mapNumber=null, fund=null), CN=ArticleExt(id=1329420412963893511, articleId=1329420410619277551, tenantId=1146029695717560320, journalId=1324759914146529478, language=CN, title=SpikingReLU:一种基于混合模型转换的高性能脉冲神经网络, columnId=1325489892104483745, journalTitle=计算机研究与发展, columnName=人工智能, runingTitle=null, highlight=null, articleAbstract=

受大脑启发的脉冲神经网络(spiking neural networks,SNNs)由于其强大的时空编码能力和事件驱动计算等特性展现出巨大的发展潜力。目前,主流的SNNs方法分为将训练完备的人工神经网络(artificial neural networks,ANNs)转换为SNNs以及直接训练SNNs这2类。然而,现有的ANNs转换方法需要较长的时间步减少转换误差,直接训练的SNNs则受限于网络表征能力的不足。针对上述挑战,提出一种基于混合模型转换的高性能脉冲神经网络。所提方法将模型训练与推理解耦:在训练阶段,策略性地将SNNs中部分脉冲神经元替换为ReLU激活函数,增强网络的特征学习能力,从而构建一个同时包含ANNs与SNNs的高性能混合模型;由于混合模型中的ANNs部分无法进行事件驱动计算,进一步引入重参数化技巧,在不损失性能的前提下,将混合模型的推理过程全部转换为事件驱动计算,从而使所提方法兼具高性能和事件驱动计算的优势。实验结果表明:所提方法在CIFAR-10和CIFAR-100数据集上的分类准确率分别为97.31%和83.34%,在ImageNet数据集上的分类准确率为70.89%,在神经静态数据集CIFAR10-DVS的分类准确率为82.71%,均优于现有国内外最先进的方法。

, authors=王宇飞1, 赵静月2, 张歌3, 樊斌4, 杨智杰2, 万众2, 戴玉超4, 王蕾2, *, authorsList=王宇飞, 赵静月, 张歌, 樊斌, 杨智杰, 万众, 戴玉超, 王蕾, authorCompany=null, correspAuthors=王蕾, authorNote=

作者贡献声明

王宇飞负责研究方案实施、实验开展与结果分析,并完成论文撰写;赵静月参与算法实验验证,负责实验结果整理及论文内容修订;张歌负责研究思路把关、理论分析指导及论文审阅;樊斌参与理论框架论证,并对论文内容和技术表述进行修改完善;杨智杰负责实验方案设计、实验流程优化及相关验证工作;万众负责论文审校、内容修改及文字规范化;戴玉超为研究提供理论指导,并参与论文结构调整与学术内容修订;王蕾负责相关理论问题指导,并对论文内容进行审阅和完善。

1995年生。博士。主要研究方向为机器视觉、类脑计算

1994年生。博士,助理研究员。CCF会员。主要研究方向为类脑计算、类脑机器人、具身智能

1996年生。博士,助理研究员。主要研究方向为高光谱图像解译、多模态图像融合

1994年生。博士,教授,博士生导师。主要研究方向为机器视觉、计算摄影、3维重建

1995 年生。博士。主要研究方向为类脑计算、计算机体系结构

1989年生。博士,副研究员。主要研究方向为类脑计算

1982年生。博士,教授,博士生导师。CCF高级会员。主要研究方向为机器视觉、具身智能

1977年生。博士,研究员,博士生导师。CCF杰出会员。主要研究方向为计算机体系结构、智能芯片

, correspAuthorsNote=
王蕾()
, copyrightStatement=版权所有 © 2026 计算机研究与发展编辑部, copyrightOwner=null, extLink=null, articleAbsUrl=null, sourceXml=XV/Z06rRMMqlbYl8eKKnMQ==, magXml=4X5OzNNuzCV5g2EfXclpFA==, pdfUrl=null, pdf=xYqtepDScJ+YHYuwkJ/DHg==, pdfFileSize=1430323, pdfExtLink=null, richHtmlUrl=null, mobilePdfUrl=null, reviewReport=null, pdfFirstPage=null, abstractGraph=9PNXn7rMNbdvsbQOpcgRMQ==, abstractGraphContent=null, abstractVideo=null, citation=null, cebUrl=null, magXmlContent=2X/6ywHMsHNFD7swkf8ZPQ==, mapNumber=null, fund=null)}, authors=[Author(id=1329420413446238485, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, orderNo=0, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=wangyufei@qiyuanlab.com, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1329420413517541655, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, authorId=1329420413446238485, language=EN, stringName=Yufei Wang, firstName=Yufei, middleName=null, lastName=Wang, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1Qiyuan Lab, Beijing 100095, bio={"content":"

born in 1995. PhD. His main research interests include machine vision and neuromorphic computing

"}, bioImg=null, bioContent=

born in 1995. PhD. His main research interests include machine vision and neuromorphic computing

, aboutCorrespAuthor=null), CN=AuthorExt(id=1329420413580456216, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, authorId=1329420413446238485, language=CN, stringName=王宇飞, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1启元实验室 北京 100095, bio={"img":"KhPhUJkZgXzeQ+KozAYuwQ==","content":"

1995年生。博士。主要研究方向为机器视觉、类脑计算

"}, bioImg=KhPhUJkZgXzeQ+KozAYuwQ==, bioContent=

1995年生。博士。主要研究方向为机器视觉、类脑计算

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1329420413156831496, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, xref=1, ext=[AuthorCompanyExt(id=1329420413165220105, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, companyId=1329420413156831496, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1Qiyuan Lab, Beijing 100095), AuthorCompanyExt(id=1329420413173608714, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, companyId=1329420413156831496, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1启元实验室 北京 100095)])]), Author(id=1329420413655953690, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, orderNo=1, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=null, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1329420413739839772, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, authorId=1329420413655953690, language=EN, stringName=Jingyue Zhao, firstName=Jingyue, middleName=null, lastName=Zhao, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=2, address=2Institute of National Defense Science and Technology Innovation, Beijing 100091, bio={"content":"

born in 1994. PhD, assistant professor. Member of CCF. Her main research interests include neuromorphic computing, brain-inspired robotics, and embodied intelligence

"}, bioImg=null, bioContent=

born in 1994. PhD, assistant professor. Member of CCF. Her main research interests include neuromorphic computing, brain-inspired robotics, and embodied intelligence

, aboutCorrespAuthor=null), CN=AuthorExt(id=1329420413823725853, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, authorId=1329420413655953690, language=CN, stringName=赵静月, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=2, address=2军事科学院国防科技创新研究院 北京 100091, bio={"img":"yFKDKr6xdq4JOe65N7A/7w==","content":"

1994年生。博士,助理研究员。CCF会员。主要研究方向为类脑计算、类脑机器人、具身智能

"}, bioImg=yFKDKr6xdq4JOe65N7A/7w==, bioContent=

1994年生。博士,助理研究员。CCF会员。主要研究方向为类脑计算、类脑机器人、具身智能

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1329420413232328971, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, xref=2, ext=[AuthorCompanyExt(id=1329420413236523276, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, companyId=1329420413232328971, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2Institute of National Defense Science and Technology Innovation, Beijing 100091), AuthorCompanyExt(id=1329420413244911885, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, companyId=1329420413232328971, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2军事科学院国防科技创新研究院 北京 100091)])]), Author(id=1329420413899223327, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, orderNo=2, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=null, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1329420413966332193, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, authorId=1329420413899223327, language=EN, stringName=Ge Zhang, firstName=Ge, middleName=null, lastName=Zhang, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=3, address=3Beijing Institute of Tracking and Telecommunication Technology, Beijing 100094, bio={"content":"

born in 1996. PhD, assistant professor. Her main research interests include hyperspectral image interpretation and multimodal image fusion

"}, bioImg=null, bioContent=

born in 1996. PhD, assistant professor. Her main research interests include hyperspectral image interpretation and multimodal image fusion

, aboutCorrespAuthor=null), CN=AuthorExt(id=1329420414062801186, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, authorId=1329420413899223327, language=CN, stringName=张歌, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=3, address=3北京跟踪与通信技术研究所 北京 100094, bio={"img":"ktZjwGCCMGbJEGqOAwMC+g==","content":"

1996年生。博士,助理研究员。主要研究方向为高光谱图像解译、多模态图像融合

"}, bioImg=ktZjwGCCMGbJEGqOAwMC+g==, bioContent=

1996年生。博士,助理研究员。主要研究方向为高光谱图像解译、多模态图像融合

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1329420413303632142, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, xref=3, ext=[AuthorCompanyExt(id=1329420413307826447, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, companyId=1329420413303632142, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=3Beijing Institute of Tracking and Telecommunication Technology, Beijing 100094), AuthorCompanyExt(id=1329420413316215056, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, companyId=1329420413303632142, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=3北京跟踪与通信技术研究所 北京 100094)])]), Author(id=1329420414188630309, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, orderNo=3, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=null, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1329420414268322087, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, authorId=1329420414188630309, language=EN, stringName=Bin Fan, firstName=Bin, middleName=null, lastName=Fan, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=4, address=4Shaanxi Key Laboratory of Information Acquisition and Processing (Northwestern Polytechnical University), Xi’an 710129, bio={"content":"

born in 1994. PhD, professor, PhD supervisor. His main research interests include machine vision, computational photography, and 3D reconstruction

"}, bioImg=null, bioContent=

born in 1994. PhD, professor, PhD supervisor. His main research interests include machine vision, computational photography, and 3D reconstruction

, aboutCorrespAuthor=null), CN=AuthorExt(id=1329420414335430952, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, authorId=1329420414188630309, language=CN, stringName=樊斌, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=4, address=4陕西省信息获取与处理重点实验室(西北工业大学) 西安 710129, bio={"img":"0CT0S8s/+xpmgX1cS2GoNQ==","content":"

1994年生。博士,教授,博士生导师。主要研究方向为机器视觉、计算摄影、3维重建

"}, bioImg=0CT0S8s/+xpmgX1cS2GoNQ==, bioContent=

1994年生。博士,教授,博士生导师。主要研究方向为机器视觉、计算摄影、3维重建

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1329420413366546705, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, xref=4, ext=[AuthorCompanyExt(id=1329420413379129618, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, companyId=1329420413366546705, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=4Shaanxi Key Laboratory of Information Acquisition and Processing (Northwestern Polytechnical University), Xi’an 710129), AuthorCompanyExt(id=1329420413383323923, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, companyId=1329420413366546705, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=4陕西省信息获取与处理重点实验室(西北工业大学) 西安 710129)])]), Author(id=1329420414444482858, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, orderNo=4, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=null, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1329420414528368940, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, authorId=1329420414444482858, language=EN, stringName=Zhijie Yang, firstName=Zhijie, middleName=null, lastName=Yang, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=2, address=2Institute of National Defense Science and Technology Innovation, Beijing 100091, bio={"content":"

born in 1995. PhD. His main research interests include neuromorphic computing and computer architecture

"}, bioImg=null, bioContent=

born in 1995. PhD. His main research interests include neuromorphic computing and computer architecture

, aboutCorrespAuthor=null), CN=AuthorExt(id=1329420414595477805, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, authorId=1329420414444482858, language=CN, stringName=杨智杰, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=2, address=2军事科学院国防科技创新研究院 北京 100091, bio={"img":"7ewoPcccusAQ2CwmLo8OpA==","content":"

1995 年生。博士。主要研究方向为类脑计算、计算机体系结构

"}, bioImg=7ewoPcccusAQ2CwmLo8OpA==, bioContent=

1995 年生。博士。主要研究方向为类脑计算、计算机体系结构

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1329420413232328971, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, xref=2, ext=[AuthorCompanyExt(id=1329420413236523276, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, companyId=1329420413232328971, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2Institute of National Defense Science and Technology Innovation, Beijing 100091), AuthorCompanyExt(id=1329420413244911885, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, companyId=1329420413232328971, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2军事科学院国防科技创新研究院 北京 100091)])]), Author(id=1329420414658392367, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, orderNo=5, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=null, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1329420414733889841, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, authorId=1329420414658392367, language=EN, stringName=Zhong Wan, firstName=Zhong, middleName=null, lastName=Wan, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=2, address=2Institute of National Defense Science and Technology Innovation, Beijing 100091, bio={"content":"

born in 1989. PhD, associate professor. Her main research interest includes neuromorphic computing

"}, bioImg=null, bioContent=

born in 1989. PhD, associate professor. Her main research interest includes neuromorphic computing

, aboutCorrespAuthor=null), CN=AuthorExt(id=1329420414805193010, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, authorId=1329420414658392367, language=CN, stringName=万众, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=2, address=2军事科学院国防科技创新研究院 北京 100091, bio={"img":"1fyh1oBMlUaSmeI3QF5z7g==","content":"

1989年生。博士,副研究员。主要研究方向为类脑计算

"}, bioImg=1fyh1oBMlUaSmeI3QF5z7g==, bioContent=

1989年生。博士,副研究员。主要研究方向为类脑计算

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1329420413232328971, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, xref=2, ext=[AuthorCompanyExt(id=1329420413236523276, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, companyId=1329420413232328971, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2Institute of National Defense Science and Technology Innovation, Beijing 100091), AuthorCompanyExt(id=1329420413244911885, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, companyId=1329420413232328971, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2军事科学院国防科技创新研究院 北京 100091)])]), Author(id=1329420414868107572, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, orderNo=6, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=null, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1329420414939410742, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, authorId=1329420414868107572, language=EN, stringName=Yuchao Dai, firstName=Yuchao, middleName=null, lastName=Dai, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=4, address=4Shaanxi Key Laboratory of Information Acquisition and Processing (Northwestern Polytechnical University), Xi’an 710129, bio={"content":"

born in 1982. PhD, professor, PhD supervisor. Senior member of CCF. His main research interests include machine vision and embodied intelligence

"}, bioImg=null, bioContent=

born in 1982. PhD, professor, PhD supervisor. Senior member of CCF. His main research interests include machine vision and embodied intelligence

, aboutCorrespAuthor=null), CN=AuthorExt(id=1329420415010713911, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, authorId=1329420414868107572, language=CN, stringName=戴玉超, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=4, address=4陕西省信息获取与处理重点实验室(西北工业大学) 西安 710129, bio={"img":"+IQ/KZFMXiIksA2+vA8IPg==","content":"

1982年生。博士,教授,博士生导师。CCF高级会员。主要研究方向为机器视觉、具身智能

"}, bioImg=+IQ/KZFMXiIksA2+vA8IPg==, bioContent=

1982年生。博士,教授,博士生导师。CCF高级会员。主要研究方向为机器视觉、具身智能

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1329420413366546705, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, xref=4, ext=[AuthorCompanyExt(id=1329420413379129618, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, companyId=1329420413366546705, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=4Shaanxi Key Laboratory of Information Acquisition and Processing (Northwestern Polytechnical University), Xi’an 710129), AuthorCompanyExt(id=1329420413383323923, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, companyId=1329420413366546705, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=4陕西省信息获取与处理重点实验室(西北工业大学) 西安 710129)])]), Author(id=1329420415090405689, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, orderNo=7, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=leiwang@nudt.edu.cn, emailSecond=null, emailThird=null, correspondingAuthor=1, authorType=1, ext={EN=AuthorExt(id=1329420415170097467, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, authorId=1329420415090405689, language=EN, stringName=Lei Wang, firstName=Lei, middleName=null, lastName=Wang, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=2, *, address=2Institute of National Defense Science and Technology Innovation, Beijing 100091, bio={"content":"

born in 1977. PhD, professor, PhD supervisor. Distinguished member of CCF. Her main research interests include computer architecture and intelligent chips

"}, bioImg=null, bioContent=

born in 1977. PhD, professor, PhD supervisor. Distinguished member of CCF. Her main research interests include computer architecture and intelligent chips

, aboutCorrespAuthor=null), CN=AuthorExt(id=1329420415233012028, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, authorId=1329420415090405689, language=CN, stringName=王蕾, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=2, *, address=2军事科学院国防科技创新研究院 北京 100091, bio={"img":"npIuc8x7vKLsZcr8zjStkg==","content":"

1977年生。博士,研究员,博士生导师。CCF杰出会员。主要研究方向为计算机体系结构、智能芯片

"}, bioImg=npIuc8x7vKLsZcr8zjStkg==, bioContent=

1977年生。博士,研究员,博士生导师。CCF杰出会员。主要研究方向为计算机体系结构、智能芯片

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1329420413232328971, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, xref=2, ext=[AuthorCompanyExt(id=1329420413236523276, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, companyId=1329420413232328971, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2Institute of National Defense Science and Technology Innovation, Beijing 100091), AuthorCompanyExt(id=1329420413244911885, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, companyId=1329420413232328971, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2军事科学院国防科技创新研究院 北京 100091)])])], keywords=[Keyword(id=1329420415316898109, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, language=EN, orderNo=1, keyword=brain-inspired computing), Keyword(id=1329420415375618366, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, language=EN, orderNo=2, keyword=spiking neural networks), Keyword(id=1329420415425950015, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, language=EN, orderNo=3, keyword=event-driven computing), Keyword(id=1329420415488864576, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, language=EN, orderNo=4, keyword=hybrid model), Keyword(id=1329420415652442433, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, language=EN, orderNo=5, keyword=training-inference decoupling method), Keyword(id=1329420415740522818, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, language=CN, orderNo=1, keyword=类脑计算), Keyword(id=1329420415891517763, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, language=CN, orderNo=2, keyword=脉冲神经网络), Keyword(id=1329420415975403844, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, language=CN, orderNo=3, keyword=事件驱动计算), Keyword(id=1329420416042512709, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, language=CN, orderNo=4, keyword=混合模型), Keyword(id=1329420416126398790, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, language=CN, orderNo=5, keyword=训练-推理解耦方法)], refs=[Reference(id=1329420417590210903, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=1, rfOrder=0, authorNames=null, journalName=null, refType=null, unstructuredReference=黄铁军, 余肇飞, 刘怡俊. 类脑机的思想与体系结构综述[J]. 计算机研究与发展, 2019, 56(6): 1135−1148, articleTitle=null, refAbstract=null), Reference(id=1329420417669902680, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=1, rfOrder=1, authorNames=null, journalName=null, refType=null, unstructuredReference=Huang Tiejun, Yu Zhaofei, Liu Yijun. A survey of brain-inspired machine thought and architecture[J]. Journal of Computer Research and Development, 2019, 56(6): 1135−1148 (in Chinese), articleTitle=null, refAbstract=null), Reference(id=1329420417791537497, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=2, rfOrder=2, authorNames=null, journalName=null, refType=null, unstructuredReference=Guo Yufei, Huang Xuhui, Ma Zhe. Direct learning-based deep spiking neural networks: A review[J]. Frontiers in Neuroscience, 2023, 17(1): 120−795, articleTitle=null, refAbstract=null), Reference(id=1329420417888006490, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=3, rfOrder=3, authorNames=null, journalName=null, refType=null, unstructuredReference=Wu Yuchen, Feng Junwen. Development and application of artificial neural network[J]. Wireless Personal Communications, 2018, 102(2): 1645−1656, articleTitle=null, refAbstract=null), Reference(id=1329420417967698268, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=4, rfOrder=4, authorNames=null, journalName=null, refType=null, unstructuredReference=Bellec G, Salaj D, Subramoney A, et al. Long short-term memory and learning-to-learn in networks of spiking neurons[C]//Proc of the 32nd Int Conf on Neural Information Processing Systems (NeurIPS). Red Hook, NY: Curran Associates Inc, 2018: 787−797, articleTitle=null, refAbstract=null), Reference(id=1329420418059972957, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=5, rfOrder=5, authorNames=null, journalName=null, refType=null, unstructuredReference=Fang Wei, Yu Zhaofei, Chen Yanqi, et al. Incorporating learnable membrane time constant to enhance learning of spiking neural networks[C]//Proc of the 21st IEEE/CVF Int Conf on Computer Vision (ICCV). Los Alamitos, CA: IEEE Computer Society, 2021: 2661−2671, articleTitle=null, refAbstract=null), Reference(id=1329420418152247646, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=6, rfOrder=6, authorNames=null, journalName=null, refType=null, unstructuredReference=Guo Yufei, Chen Yuanpei, Liu Xiaode, et al. Ternary spike: Learning ternary spikes for spiking neural networks[C]//Proc of the 38th AAAI Conf on Artificial Intelligence. Palo Alto, CA: AAAI, 2024, 12244−12252, articleTitle=null, refAbstract=null), Reference(id=1329420418236133727, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=7, rfOrder=7, authorNames=null, journalName=null, refType=null, unstructuredReference=Luo Xinhao, Yao Man, Chou Yuhong, et al. Integer-valued training and spike-driven inference spiking neural network for high-performance and energy-efficient object detection[C]//Proc of the 18th European Conf on Computer Vision (ECCV). Berlin: Springer, 2024: 253−272, articleTitle=null, refAbstract=null), Reference(id=1329420418315825504, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=8, rfOrder=8, authorNames=null, journalName=null, refType=null, unstructuredReference=Hu Yifan, Deng Lei, Wu Yujie, et al. Advancing spiking neural networks toward deep residual learning[J]. IEEE Transactions on Neural Networks and Learning Systems, 2024, 36(2): 2353−2367, articleTitle=null, refAbstract=null), Reference(id=1329420418374545761, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=9, rfOrder=9, authorNames=null, journalName=null, refType=null, unstructuredReference=Su Qiaoyi, Chou Yuhong, Hu Yifan, et al. Deep directly-trained spiking neural networks for object detection[C]//Proc of the 22nd IEEE/CVF Int Conf on Computer Vision (ICCV). Los Alamitos, CA: IEEE Computer Society, 2023: 6555−6565, articleTitle=null, refAbstract=null), Reference(id=1329420418429071714, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=10, rfOrder=10, authorNames=null, journalName=null, refType=null, unstructuredReference=Yao Man, Zhao Guangshe, Zhang Hengyu, et al. Attention spiking neural networks[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2023, 45(8): 9393−9410, articleTitle=null, refAbstract=null), Reference(id=1329420418500374883, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=11, rfOrder=11, authorNames=null, journalName=null, refType=null, unstructuredReference=Yang Shu, Yu Chengting, Liu Lei, et al. Efficient ANN-guided distillation: Aligning rate-based features of spiking neural networks through hybrid block-wise replacement[C]//Proc of the 42nd IEEE/CVF Conf on Computer Vision and Pattern Recognition (CVPR). Los Alamitos, CA: IEEE Computer Society, 2025: 10025−10035, articleTitle=null, refAbstract=null), Reference(id=1329420418567483748, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=12, rfOrder=12, authorNames=null, journalName=null, refType=null, unstructuredReference=Yu Chengting, Zhao Xiaochen, Liu Lei, et al. Efficient logit-based knowledge distillation of deep spiking neural networks for full-range timestep deployment[C/OL]//Proc of the 42nd Int Conf on Machine Learning (ICML). 2025: 393−941. [2026-01-21]. https://openreview.net/forum?id=ZvkyeUrpsA, articleTitle=null, refAbstract=null), Reference(id=1329420418655564133, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=13, rfOrder=13, authorNames=null, journalName=null, refType=null, unstructuredReference=Cao Yongqiang, Chen Yang, Khosla D. Spiking deep convolutional neural networks for energy-efficient object recognition[J]. International Journal of Computer Vision, 2015, 113(1): 54−66, articleTitle=null, refAbstract=null), Reference(id=1329420418731061606, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=14, rfOrder=14, authorNames=null, journalName=null, refType=null, unstructuredReference=Hunsberger E, Eliasmith C. Spiking deep networks with LIF neurons[J]. arXiv preprint, arXiv: 1510.08829, 2015, articleTitle=null, refAbstract=null), Reference(id=1329420418793976167, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=15, rfOrder=15, authorNames=null, journalName=null, refType=null, unstructuredReference=Hunsberger E, Eliasmith C. Training spiking deep networks for neuromorphic hardware[J]. arXiv preprint, arXiv: 1611.05141, 2016, articleTitle=null, refAbstract=null), Reference(id=1329420418869473640, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=16, rfOrder=16, authorNames=null, journalName=null, refType=null, unstructuredReference=Diehl P U, Neil D, Binas J, et al. Fast-classifying, high-accuracy spiking deep networks through weight and threshold balancing[C]//Proc of the 2015 Int Joint Conf on Neural Networks (IJCNN). Piscataway, NJ: IEEE, 2015: 1−8, articleTitle=null, refAbstract=null), Reference(id=1329420418961748329, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=17, rfOrder=17, authorNames=null, journalName=null, refType=null, unstructuredReference=Zhang Anguo, Shi Jieming, Wu Junyi, et al. Low latency and sparse computing spiking neural networks with self-driven adaptive threshold plasticity[J]. IEEE Transactions on Neural Networks and Learning Systems, 2024, 35(12): 17177−17188, articleTitle=null, refAbstract=null), Reference(id=1329420419028857194, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=18, rfOrder=18, authorNames=null, journalName=null, refType=null, unstructuredReference=Bu Tong, Ding Jianhao, Yu Zhaofei, et al. Optimized potential initialization for low-latency spiking neural networks[C]//Proc of the 36th AAAI Conf on Artificial Intelligence. Palo Alto, CA: AAAI, 2022, 11−20, articleTitle=null, refAbstract=null), Reference(id=1329420419091771755, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=19, rfOrder=19, authorNames=null, journalName=null, refType=null, unstructuredReference=Yao Xingting, Li Fanrong, Mo Zitao, et al. GLIF: A unified gated leaky integrate-and-fire neuron for spiking neural networks[C]//Proc of the 36th Int Conf on Neural Information Processing Systems (NeurIPS). San Diego, CA: Neural Information Processing Systems Foundation, Inc, 2022: 32160−32171, articleTitle=null, refAbstract=null), Reference(id=1329420419184046444, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=20, rfOrder=20, authorNames=null, journalName=null, refType=null, unstructuredReference=Guo Yufei, Liu Xiaode, Chen Yuanpei, et al. Spiking Transformer: introducing accurate addition-only spiking self-attention for transformer[C]//Proc of the 42nd IEEE/CVF Conf on Computer Vision and Pattern Recognition (CVPR). Los Alamitos, CA: IEEE Computer Society, 2025: 24398−24408, articleTitle=null, refAbstract=null), Reference(id=1329420419259543917, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=21, rfOrder=21, authorNames=null, journalName=null, refType=null, unstructuredReference=Guo Yufei, Zhang Liwen, Chen Yuanpei, et al. Real spike: Learning real-valued spikes for spiking neural networks[C]//Proc of the 17th European Conf on Computer Vision (ECCV). Berlin: Springer, 2022: 52−68, articleTitle=null, refAbstract=null), Reference(id=1329420419335041390, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=22, rfOrder=22, authorNames=null, journalName=null, refType=null, unstructuredReference=冯烁, 路冬冬, 尹飞, 等. 面向分布式计算的类脑智能处理器指令集设计[J]. 计算机研究与发展, 2025, 62(1): 1−14, articleTitle=null, refAbstract=null), Reference(id=1329420419410538863, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=22, rfOrder=23, authorNames=null, journalName=null, refType=null, unstructuredReference=Feng Shuo, Lu Dongdong, Yin Fei, et al. Instruction set design for brain-inspired intelligent processors oriented to distributed computing[J]. Journal of Computer Research and Development, 2025, 62(1): 1−14 (in Chinese), articleTitle=null, refAbstract=null), Reference(id=1329420419490230640, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=23, rfOrder=24, authorNames=null, journalName=null, refType=null, unstructuredReference=杨智杰, 王蕾, 石伟, 等. 类脑处理器异步片上网络架构[J]. 计算机研究与发展, 2023, 60(1): 17−29, articleTitle=null, refAbstract=null), Reference(id=1329420419578311025, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=23, rfOrder=25, authorNames=null, journalName=null, refType=null, unstructuredReference=Yang Zhijie, Wang Lei, Shi Wei, et al. Asynchronous network-on-chip architecture for brain-inspired processors[J]. Journal of Computer Research and Development, 2023, 60(1): 17−29 (in Chinese), articleTitle=null, refAbstract=null), Reference(id=1329420419641225586, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=24, rfOrder=26, authorNames=null, journalName=null, refType=null, unstructuredReference=王智超, 陈亮, 李千鹏, 等. 基于混合模式路由的脑启发片上网络架构[J]. 计算机研究与发展, 2025, 62(11): 2663−2676, articleTitle=null, refAbstract=null), Reference(id=1329420419725111667, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=24, rfOrder=27, authorNames=null, journalName=null, refType=null, unstructuredReference=Wang Zhichao, Chen Liang, Li Qianpeng, et al. A brain-inspired network-on-chip architecture with hybrid-mode routing[J]. Journal of Computer Research and Development, 2025, 62(11): 2663−2676 (in Chinese), articleTitle=null, refAbstract=null), Reference(id=1329420419821580660, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=25, rfOrder=28, authorNames=null, journalName=null, refType=null, unstructuredReference=Sengupta A, Ye Yuting, Wang R, et al. Going deeper in spiking neural networks: VGG and residual architectures[J]. Frontiers in Neuroscience, 2019, 13(1): 95−106, articleTitle=null, refAbstract=null), Reference(id=1329420419888689525, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=26, rfOrder=29, authorNames=null, journalName=null, refType=null, unstructuredReference=Han Bing, Srinivasan G, Roy K. RMP-SNN: Residual membrane potential neuron for enabling deeper high-accuracy and low-latency spiking neural network[C]//Proc of the 37th IEEE/CVF Conf on Computer Vision and Pattern Recognition (CVPR). Los Alamitos, CA: IEEE Computer Society, 2020: 13558−13567, articleTitle=null, refAbstract=null), Reference(id=1329420419972575606, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=27, rfOrder=30, authorNames=null, journalName=null, refType=null, unstructuredReference=Yoo D, Jeong D S. Cbp-Qsnn: Spiking neural networks quantized using constrained backpropagation[J]. IEEE Journal on Emerging and Selected Topics in Circuits and Systems, 2023, 13(4): 1137−1146, articleTitle=null, refAbstract=null), Reference(id=1329420420060655991, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=28, rfOrder=31, authorNames=null, journalName=null, refType=null, unstructuredReference=Meng Qingyan, Xiao Mingqing, Yan Shen, et al. Training high-performance low-latency spiking neural networks by differentiation on spike representation[C]//Proc of the 39th IEEE/CVF Conf on Computer Vision and Pattern Recognition (CVPR). Los Alamitos, CA: IEEE Computer Society, 2022: 12444−12453, articleTitle=null, refAbstract=null), Reference(id=1329420420136153464, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=29, rfOrder=32, authorNames=null, journalName=null, refType=null, unstructuredReference=Zhou Zhaokun, Zhu Yuesheng, He Chao, et al. SpikFormer: When spiking neural network meets transformer[C/OL]//Proc of the 11th Int Conf on Learning Representations (ICLR). 2023[2026-01-21]. https://openreview.net/forum?id=frE4fUwz_h, articleTitle=null, refAbstract=null), Reference(id=1329420420215845241, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=30, rfOrder=33, authorNames=null, journalName=null, refType=null, unstructuredReference=Yao Man, Hu Jiakui, Zhou Zhaokun, et al. Spike-driven Transformer[C]//Proc of the 37th Int Conf on Neural Information Processing Systems (NeurIPS). New Orleans, CA: Curran Associates, 2023: 64043−64058, articleTitle=null, refAbstract=null), Reference(id=1329420420303925626, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=31, rfOrder=34, authorNames=null, journalName=null, refType=null, unstructuredReference=Fang Yi, Wang Zhaokun, Zhang Lei, et al. Spiking wavelet Transformer[C]//Proc of the 18th European Conf on Computer Vision (ECCV). Berlin: Springer, 2024: 19−37, articleTitle=null, refAbstract=null), Reference(id=1329420420379423099, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=32, rfOrder=35, authorNames=null, journalName=null, refType=null, unstructuredReference=Wang Shu, Zhang Ming, Zhang Di, et al. Spiking vision Transformer with saccadic attention[C/OL]//Proc of the Int Conf on Learning Representations (ICLR). 2025[2026-01-21]. https://openreview.net/forum?id=qzZsz6MuEq, articleTitle=null, refAbstract=null), Reference(id=1329420420467503484, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=33, rfOrder=36, authorNames=null, journalName=null, refType=null, unstructuredReference=Liang Yixing, Wei Wenjie, Belatreche A, et al. Towards accurate binary spiking neural networks: Learning with adaptive gradient modulation mechanism[C]//Proc of the 39th AAAI Conf on Artificial Intelligence. Palo Alto, CA: AAAI, 2025, 39(2): 1402−1410, articleTitle=null, refAbstract=null), Reference(id=1329420420534612349, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=34, rfOrder=37, authorNames=null, journalName=null, refType=null, unstructuredReference=Zhang Tianpeng, Yu Kexin, Zhong Xinghao, et al. STAA-SNN: Spatial-temporal attention aggregator for spiking neural networks[C]//Proc of the 42nd IEEE/CVF Conf on Computer Vision and Pattern Recognition (CVPR). Los Alamitos, CA: IEEE Computer Society, 2025: 13959−13969, articleTitle=null, refAbstract=null), Reference(id=1329420420618498430, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=35, rfOrder=38, authorNames=null, journalName=null, refType=null, unstructuredReference=Deng Shikuang, Li Yang, Zhang Shiyan, et al. Temporal efficient training of spiking neural network via gradient re-weighting[C/OL]//Proc of the Int Conf on Learning Representations (ICLR). 2022[2026-01-21]. https://openreview.net/forum?id=_XNtisL32jv, articleTitle=null, refAbstract=null), Reference(id=1329420420685607295, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=36, rfOrder=39, authorNames=null, journalName=null, refType=null, unstructuredReference=Guo Yufei, Tong Xinyi, Chen Yuanpei, et al. RecDis-SNN: Rectifying membrane potential distribution for directly training spiking neural networks[C]//Proc of the 39th IEEE/CVF Conf on Computer Vision and Pattern Recognition (CVPR). Los Alamitos, CA: IEEE Computer Society, 2022: 326−335, articleTitle=null, refAbstract=null), Reference(id=1329420420752716160, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=37, rfOrder=40, authorNames=null, journalName=null, refType=null, unstructuredReference=Guo Yufei, Liu Xiaode, Chen Yuanpei, et al. RMP-loss: Regularizing membrane potential distribution for spiking neural networks[C]//Proc of the 22nd IEEE/CVF Int Conf on Computer Vision (ICCV). Los Alamitos, CA: IEEE Computer Society, 2023: 17391−17401, articleTitle=null, refAbstract=null), Reference(id=1329420420828213633, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=38, rfOrder=41, authorNames=null, journalName=null, refType=null, unstructuredReference=Guo Yufei, Zhang Yuhan, Chen Yuanpei, et al. Membrane potential batch normalization for spiking neural networks[C]//Proc of the 22nd IEEE/CVF Int Conf on Computer Vision (ICCV). Los Alamitos, CA: IEEE Computer Society, 2023: 19420−19430, articleTitle=null, refAbstract=null), Reference(id=1329420420937265538, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=39, rfOrder=42, authorNames=null, journalName=null, refType=null, unstructuredReference=Yu Kexin, Zhang Tianpeng, Wang Hao, et al. FSTA-SNN: Frequency-based spatial-temporal attention module for spiking neural networks[C]//Proc of the 39th AAAI Conf on Artificial Intelligence. Palo Alto, CA: AAAI, 2025: 22227−22235, articleTitle=null, refAbstract=null), Reference(id=1329420421054706051, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=40, rfOrder=43, authorNames=null, journalName=null, refType=null, unstructuredReference=Guo Yufei, Chen Yuanpei, Zhang Liwen, et al. IM-loss: Information maximization loss for spiking neural networks[C]//Proc of the 36th Int Conf on Neural Information Processing Systems (NeurIPS). San Diego, CA: Neural Information Processing Systems Foundation, Inc, 2022: 156−166, articleTitle=null, refAbstract=null)], funds=null, companyList=[AuthorCompany(id=1329420413156831496, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, xref=1, ext=[AuthorCompanyExt(id=1329420413165220105, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, companyId=1329420413156831496, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1Qiyuan Lab, Beijing 100095), AuthorCompanyExt(id=1329420413173608714, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, companyId=1329420413156831496, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1启元实验室 北京 100095)]), AuthorCompany(id=1329420413232328971, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, xref=2, ext=[AuthorCompanyExt(id=1329420413236523276, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, companyId=1329420413232328971, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2Institute of National Defense Science and Technology Innovation, Beijing 100091), AuthorCompanyExt(id=1329420413244911885, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, companyId=1329420413232328971, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2军事科学院国防科技创新研究院 北京 100091)]), AuthorCompany(id=1329420413303632142, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, xref=3, ext=[AuthorCompanyExt(id=1329420413307826447, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, companyId=1329420413303632142, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=3Beijing Institute of Tracking and Telecommunication Technology, Beijing 100094), AuthorCompanyExt(id=1329420413316215056, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, companyId=1329420413303632142, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=3北京跟踪与通信技术研究所 北京 100094)]), AuthorCompany(id=1329420413366546705, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, xref=4, ext=[AuthorCompanyExt(id=1329420413379129618, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, companyId=1329420413366546705, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=4Shaanxi Key Laboratory of Information Acquisition and Processing (Northwestern Polytechnical University), Xi’an 710129), AuthorCompanyExt(id=1329420413383323923, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, companyId=1329420413366546705, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=4陕西省信息获取与处理重点实验室(西北工业大学) 西安 710129)])], figs=[ArticleFig(id=1329420416289976647, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, language=EN, label=Figure 1, caption=The proposed SpikingReLU module and traditional CBS module, figureFileSmall=niM90ZzvQKjhirx8SoNCog==, figureFileBig=9PNXn7rMNbdvsbQOpcgRMQ==, tableContent=null), ArticleFig(id=1329420416369668424, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, language=CN, label=图1, caption=本文提出的SpikingReLU模型与传统CBS模块, figureFileSmall=niM90ZzvQKjhirx8SoNCog==, figureFileBig=9PNXn7rMNbdvsbQOpcgRMQ==, tableContent=null), ArticleFig(id=1329420416545829193, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, language=EN, label=Figure 2, caption=The schematic diagram of hybrid model conversion based on the training-inference decoupling strategy, figureFileSmall=q+GgYJ3xKatLtY/Hun5ilQ==, figureFileBig=XZCjJ2G7F08vvjp6AEJ6sw==, tableContent=null), ArticleFig(id=1329420416608743754, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, language=CN, label=图2, caption=基于训练-推理解耦策略的混合模型转换示意图, figureFileSmall=q+GgYJ3xKatLtY/Hun5ilQ==, figureFileBig=XZCjJ2G7F08vvjp6AEJ6sw==, tableContent=null), ArticleFig(id=1329420416667464011, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, language=EN, label=Figure 3, caption=Feature visualization diagrams at different time steps, figureFileSmall=u+sauvxxpzNv2LaLcogSSQ==, figureFileBig=v7VRNjRiO3ri+oF761MPGg==, tableContent=null), ArticleFig(id=1329420416738767180, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, language=CN, label=图3, caption=不同时间步下的特征可视化图, figureFileSmall=u+sauvxxpzNv2LaLcogSSQ==, figureFileBig=v7VRNjRiO3ri+oF761MPGg==, tableContent=null), ArticleFig(id=1329420416801681741, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, language=EN, label=Table 1, caption=

Ablation Studies for Accuracy Results on CIFAR-10 Dataset %

, figureFileSmall=null, figureFileBig=null, tableContent=
模型神经元模块的数目
123
CBS模块2值95.8296.1996.25
Real spike[21]95.8596.3196.34
SpikingReLU2值95.8297.0597.31
3值97.47
4值97.46
), ArticleFig(id=1329420416881373518, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, language=CN, label=表1, caption=

在CIFAR-10数据集上的消融实验准确率结果

, figureFileSmall=null, figureFileBig=null, tableContent=
模型神经元模块的数目
123
CBS模块2值95.8296.1996.25
Real spike[21]95.8596.3196.34
SpikingReLU2值95.8297.0597.31
3值97.47
4值97.46
), ArticleFig(id=1329420416952676687, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, language=EN, label=Table 2, caption=

Performance Results Compared with State-of-the-Art Methods on CIFAR-10 and CIFAR-100 Datasets

, figureFileSmall=null, figureFileBig=null, tableContent=
类型方法时间步准确率/%(CIFAR-10)准确率/%(CIFAR-100)发表渠道
ANNs转换方法SpikeNorm[25]250091.55Front. Neurosci
RMP[26]204867.82CVPR
Cbp-Qsnn[27]3291.5166.63JETCAS
直接训练SNNsReal spike[21]1095.7871.24ECCV
DSR[28]2095.4078.50CVPR
SpikeFormer[29]495.5178.21ICLR
SD Transformer[30]495.6078.40NeurIPS
Ternary Spike[6]294.4873.41AAAI
494.9674.02
SWformer[31]696.3079.60ECCV
SNN-Vit[32]96.1080.10ICLR
AGMM[33]296.1380.25AAAI
496.3380.70
Spiking Transformer[20]496.4279.90CVPR
STAA-SNN[34]196.7579.37CVPR
296.8580.57
497.1482.05
混合模型转换SpikingReLU(本文)196.7582.42
297.1583.03
497.3183.34
), ArticleFig(id=1329420417049145680, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, language=CN, label=表2, caption=

在CIFAR-10和CIFAR-100数据集上与最先进方法的性能对比结果

, figureFileSmall=null, figureFileBig=null, tableContent=
类型方法时间步准确率/%(CIFAR-10)准确率/%(CIFAR-100)发表渠道
ANNs转换方法SpikeNorm[25]250091.55Front. Neurosci
RMP[26]204867.82CVPR
Cbp-Qsnn[27]3291.5166.63JETCAS
直接训练SNNsReal spike[21]1095.7871.24ECCV
DSR[28]2095.4078.50CVPR
SpikeFormer[29]495.5178.21ICLR
SD Transformer[30]495.6078.40NeurIPS
Ternary Spike[6]294.4873.41AAAI
494.9674.02
SWformer[31]696.3079.60ECCV
SNN-Vit[32]96.1080.10ICLR
AGMM[33]296.1380.25AAAI
496.3380.70
Spiking Transformer[20]496.4279.90CVPR
STAA-SNN[34]196.7579.37CVPR
296.8580.57
497.1482.05
混合模型转换SpikingReLU(本文)196.7582.42
297.1583.03
497.3183.34
), ArticleFig(id=1329420417145614673, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, language=EN, label=Table 3, caption=

Comparative Results for Classification Accuracy on ImageNet Dataset

, figureFileSmall=null, figureFileBig=null, tableContent=
方法时间步准确率/%
TET[35]664.79
RecDis-SNN[36]667.33
GLIF[19]467.52
Real spike[21]467.69
RMP-loss[37]465.17
MPBN[38]464.71
Ternary Spike[6]470.74
AGMM[33]464.67
STAA-SNN[34]470.04
FATA-SNN[39]470.23
SpikingReLU(本文)470.89
), ArticleFig(id=1329420417221112146, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, language=CN, label=表3, caption=

在ImageNet数据集上的分类准确率对比结果

, figureFileSmall=null, figureFileBig=null, tableContent=
方法时间步准确率/%
TET[35]664.79
RecDis-SNN[36]667.33
GLIF[19]467.52
Real spike[21]467.69
RMP-loss[37]465.17
MPBN[38]464.71
Ternary Spike[6]470.74
AGMM[33]464.67
STAA-SNN[34]470.04
FATA-SNN[39]470.23
SpikingReLU(本文)470.89
), ArticleFig(id=1329420417296609619, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, language=EN, label=Table 4, caption=

Comparative Results for Classification Accuracy on CIFAR10-DVS Dataset

, figureFileSmall=null, figureFileBig=null, tableContent=
方法时间步准确率/%
IM-loss[40]1072.60
SpikeFormer[29]1680.90
MPBN[38]1078.70
Ternary spike[6]1079.80
AGMM[33]482.40
STAA-SNN[34]1682.10
FATA-SNN[39]1682.70
SpikingReLU(本文)1682.71
), ArticleFig(id=1329420417359524180, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, language=CN, label=表4, caption=

在CIFAR10-DVS数据集上的分类准确率对比结果

, figureFileSmall=null, figureFileBig=null, tableContent=
方法时间步准确率/%
IM-loss[40]1072.60
SpikeFormer[29]1680.90
MPBN[38]1078.70
Ternary spike[6]1079.80
AGMM[33]482.40
STAA-SNN[34]1682.10
FATA-SNN[39]1682.70
SpikingReLU(本文)1682.71
), ArticleFig(id=1329420417430827349, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, language=EN, label=Table 5, caption=

Power Comparison on CIFAR10 Dataset mJ

, figureFileSmall=null, figureFileBig=null, tableContent=
方法功耗
ANN239.2
SNN(本文)87.3
), ArticleFig(id=1329420417493741910, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, language=CN, label=表5, caption=

在CIFAR10数据集上的功耗对比

, figureFileSmall=null, figureFileBig=null, tableContent=
方法功耗
ANN239.2
SNN(本文)87.3
)], attaches=null, journal=Journal(id=1324759606947315908, delFlag=0, nameCn=计算机研究与发展, nameEn=Journal of Computer Research and Development, nameHistory1=null, nameHistory2=null, issn=1000-1239, eissn=null, cn=11-1777/TP, coden=null, periodic=0, language=CN, oaType=null, ccby=null, superviseOffice=null, ownerOffice=null, pubOffice=null, editorOffice=null, officeType=null, aims=null, clcCode=null, officeProv=null, officeCity=null, officeAddr=null, officeZip=null, officeEmail=null, officePhone=null, editDirector=null, officeDirector=null, officeDirectorPhone=null, officeStaffNum=null, officeEmpNum=null, coverPicUrl=k2l8zgqwstuxxevFX2qxVA==, journalPrice=null, startedYear=null, abbrevIsoEn=Journal of Computer Research and Development, journalRemark=null, publicationField=null, createdTime=1793776903141, updatedTime=1793850698612, createdBy=18614031015, updatedBy=13701087609, firstLetterCn=J, firstLetterEn=J, subjectCode=Electronic Information, subjectName=null, subjectCodeEn=Electronic Information, subjectNameEn=null, picCn=k2l8zgqwstuxxevFX2qxVA==, picEn=r6k0Ry+RD+c/VXj5QarmDw==, jcr=null, cjcr=null, exts=[JournalExt(id=1325069127710286688, language=CN, name=计算机研究与发展, nameHistory1=null, nameHistory2=null, managedBy=, sponsoredBy=, publishedBy=, editorOffice=, officeProv=null, officeCity=null, officeAddr=, officeZip=, editDirector=, officeDirector=null, officePhone=null, coverPicUrl=null, journalRemark=, submitArticleUrl=null, websiteUrl=, createdTime=1793850698640, updatedTime=1793850698640, createdBy=13701087609, updatedBy=13701087609, submissionGuidelinesUrl=, submissionAuthorUrl=https://crad.ict.ac.cn/journalx/authorLogOn.action, submissionEditorUrl=https://crad.ict.ac.cn/journalx/editorLogOn.action, submissionReviewUrl=https://crad.ict.ac.cn/journalx/expertLogOn.action, submissionCeEditorUrl=https://crad.ict.ac.cn/journalx/editorInChiefLogOn.action, submissionAeEditorUrl=, option={"copyright":""}), JournalExt(id=1325069127777395553, language=EN, name=Journal of Computer Research and Development, nameHistory1=null, nameHistory2=null, managedBy=, sponsoredBy=, publishedBy=, editorOffice=, officeProv=null, officeCity=null, officeAddr=, officeZip=, editDirector=, officeDirector=null, officePhone=null, coverPicUrl=null, journalRemark=, submitArticleUrl=null, websiteUrl=, createdTime=1793850698656, updatedTime=1793850698656, createdBy=13701087609, updatedBy=13701087609, submissionGuidelinesUrl=, submissionAuthorUrl=https://crad.ict.ac.cn/journalx/authorLogOn.action, submissionEditorUrl=https://crad.ict.ac.cn/journalx/editorLogOn.action, submissionReviewUrl=https://crad.ict.ac.cn/journalx/expertLogOn.action, submissionCeEditorUrl=https://crad.ict.ac.cn/journalx/editorInChiefLogOn.action, submissionAeEditorUrl=, option={"copyright":""})], databaseList=null, tenantJournalId=1324759914146529478, websiteList=[Website(id=1325070300139892771, webName=null, webTitle=null, webDomain=null, webCopyrigh=null, webIpcNo=null, seoTitle=null, seoKeywords=null, seoDescription=null, tenantJournalId=null, journalId=1324759914146529478, journalNameCn=null, journalNameEn=null, grayFlag=null, tenantId=1146029695717560320, platformId=null, journalGroupId=null, journalGroupNameCn=null, journalGroupNameEn=null, type=1, domain=https://castjournals.cast.org.cn/joweb/jsjyjyfz/CN, language=CN, createTime=1793850978169, createBy=18614031015, updateTime=1793851040259, updateBy=18614031015, name=计算机研究与发展-中文, tplId=1146099689490845704, title=计算机研究与发展, delFlag=0, indexPage=/home, props=[WebsiteProps(id=1325070893751349360, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1325070300139892771, code=articleTextType, value=kx, createTime=1793851119697, updateTime=1793851119697, creator=18614031015, updator=18614031015), WebsiteProps(id=1325070893730377837, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1325070300139892771, code=banner, value=null, createTime=1793851119692, updateTime=1793851119692, creator=18614031015, updator=18614031015), WebsiteProps(id=1325070893772320883, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1325070300139892771, code=grayFlag, value=0, createTime=1793851119702, updateTime=1793851119702, creator=18614031015, updator=18614031015), WebsiteProps(id=1325070893721989228, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1325070300139892771, code=logo, value=https://castjournals.cast.org.cn/joweb/jsjyjyfz/CN/file/pic?fileId=U27tDsy717YF7XSPA+Vwdw==, createTime=1793851119690, updateTime=1793851119690, creator=18614031015, updator=18614031015), WebsiteProps(id=1325070893784903797, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1325070300139892771, code=minRunFlag, value=0, createTime=1793851119705, updateTime=1793851119705, creator=18614031015, updator=18614031015), WebsiteProps(id=1325070893742960751, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1325070300139892771, code=picServerUrl, value=https://castjournals.cast.org.cn/joweb/jsjyjyfz/CN/file/pic, createTime=1793851119695, updateTime=1793851119695, creator=18614031015, updator=18614031015), WebsiteProps(id=1325070893780709492, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1325070300139892771, code=silenceFlag, value=0, createTime=1793851119704, updateTime=1793851119704, creator=18614031015, updator=18614031015), WebsiteProps(id=1325070893738766446, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1325070300139892771, code=staticResourcePath, value=https://castjournals.cast.org.cn/joweb/cast_kjdb_cn_619/, createTime=1793851119694, updateTime=1793851119694, creator=18614031015, updator=18614031015), WebsiteProps(id=1325070893755543665, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1325070300139892771, code=themeColor, value=null, createTime=1793851119698, updateTime=1793851119698, creator=18614031015, updator=18614031015), WebsiteProps(id=1325070893763932274, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1325070300139892771, code=themeStyle, value=null, createTime=1793851119700, updateTime=1793851119700, creator=18614031015, updator=18614031015)]), Website(id=1325070300580294698, webName=null, webTitle=null, webDomain=null, webCopyrigh=null, webIpcNo=null, seoTitle=null, seoKeywords=null, seoDescription=null, tenantJournalId=null, journalId=1324759914146529478, journalNameCn=null, journalNameEn=null, grayFlag=null, tenantId=1146029695717560320, platformId=null, journalGroupId=null, journalGroupNameCn=null, journalGroupNameEn=null, type=1, domain=https://castjournals.cast.org.cn/joweb/jsjyjyfz/EN, language=EN, createTime=1793850978274, createBy=18614031015, updateTime=1793851036638, updateBy=18614031015, name=计算机研究与发展-英文, tplId=1146101810881728533, title=Journal of Computer Research and Development, delFlag=0, indexPage=/home, props=[WebsiteProps(id=1325070925288321148, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1325070300580294698, code=articleTextType, value=kx, createTime=1793851127216, updateTime=1793851127216, creator=18614031015, updator=18614031015), WebsiteProps(id=1325070925263155321, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1325070300580294698, code=banner, value=null, createTime=1793851127210, updateTime=1793851127210, creator=18614031015, updator=18614031015), WebsiteProps(id=1325070925309292671, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1325070300580294698, code=grayFlag, value=0, createTime=1793851127221, updateTime=1793851127221, creator=18614031015, updator=18614031015), WebsiteProps(id=1325070925254766712, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1325070300580294698, code=logo, value=https://castjournals.cast.org.cn/joweb/jsjyjyfz/EN/file/pic?fileId=U27tDsy717YF7XSPA+Vwdw==, createTime=1793851127208, updateTime=1793851127208, creator=18614031015, updator=18614031015), WebsiteProps(id=1325070925326069889, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1325070300580294698, code=minRunFlag, value=0, createTime=1793851127225, updateTime=1793851127225, creator=18614031015, updator=18614031015), WebsiteProps(id=1325070925279932539, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1325070300580294698, code=picServerUrl, value=https://castjournals.cast.org.cn/joweb/jsjyjyfz/EN/file/pic, createTime=1793851127214, updateTime=1793851127214, creator=18614031015, updator=18614031015), WebsiteProps(id=1325070925317681280, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1325070300580294698, code=silenceFlag, value=0, createTime=1793851127223, updateTime=1793851127223, creator=18614031015, updator=18614031015), WebsiteProps(id=1325070925271543930, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1325070300580294698, code=staticResourcePath, value=https://castjournals.cast.org.cn/joweb/cast_kjdb_en_623/, createTime=1793851127212, updateTime=1793851127212, creator=18614031015, updator=18614031015), WebsiteProps(id=1325070925292515453, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1325070300580294698, code=themeColor, value=null, createTime=1793851127217, updateTime=1793851127217, creator=18614031015, updator=18614031015), WebsiteProps(id=1325070925300904062, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1325070300580294698, code=themeStyle, value=null, createTime=1793851127219, updateTime=1793851127219, creator=18614031015, updator=18614031015)])], journalTitle=计算机研究与发展, weixinUrl=null, journalUrl=https://crad.ict.ac.cn/, iacademicId=null, status=1, seqNo=null, journalTitleEn=Journal of Computer Research and Development, journalPhotoCn=k2l8zgqwstuxxevFX2qxVA==, journalPhotoEn=r6k0Ry+RD+c/VXj5QarmDw==, journalFirstLetter=J, journalRecommend=null, journalNew=null, journalCollection=null, jcrJf=null, cjcrJf=null, jcrJfStr=null, cjcrJfStr=null, submissionFirstDecision=null, sciSubjectClassification=null, casSubjectClassification=null, citeScore=null, totalCitationFrequency=null, icpCode=null, psCode=null, advertisingLicenseCode=null, copyrightInformation=null, country=null, option=, provinceCode=null, provinceName=null, collectFlag=false, interPubPlatform=, interPubPlatformUrl=null), detailUrlCn=https://castjournals.cast.org.cn/joweb/jsjyjyfz/CN/10.7544/issn1000-1239.202550722, detailUrlEn=https://castjournals.cast.org.cn/joweb/jsjyjyfz/EN/10.7544/issn1000-1239.202550722, pdfUrlCn=https://castjournals.cast.org.cn/joweb/jsjyjyfz/CN/PDF/10.7544/issn1000-1239.202550722, pdfUrlEn=https://castjournals.cast.org.cn/joweb/jsjyjyfz/EN/PDF/10.7544/issn1000-1239.202550722, aliStartDate=0, aliEndDate=0, collectionFlag=false, citedCount=null, citedUrl=null, previewStatus=0, delFlag=0, hasFullText=1, orderTime=1785513600000, fullTextJson=null, articleText=null, reference=null)
收藏切换
SpikingReLU:一种基于混合模型转换的高性能脉冲神经网络
王宇飞 1 , 赵静月 2 , 张歌 3 , 樊斌 4 , 杨智杰 2 , 万众 2 , 戴玉超 4 , 王蕾 2, *
计算机研究与发展 | 人工智能 2026,63(8): 2143-2153
收起
收藏切换
计算机研究与发展 |人工智能 2026 , 63 (8) : 2143 -2153
SpikingReLU:一种基于混合模型转换的高性能脉冲神经网络
全屏
[Author(id=1329420413446238485, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, orderNo=0, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=wangyufei@qiyuanlab.com, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1329420413517541655, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, authorId=1329420413446238485, language=EN, stringName=Yufei Wang, firstName=Yufei, middleName=null, lastName=Wang, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1Qiyuan Lab, Beijing 100095, bio={"content":"

born in 1995. PhD. His main research interests include machine vision and neuromorphic computing

"}, bioImg=null, bioContent=

born in 1995. PhD. His main research interests include machine vision and neuromorphic computing

, aboutCorrespAuthor=null), CN=AuthorExt(id=1329420413580456216, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, authorId=1329420413446238485, language=CN, stringName=王宇飞, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1启元实验室 北京 100095, bio={"img":"KhPhUJkZgXzeQ+KozAYuwQ==","content":"

1995年生。博士。主要研究方向为机器视觉、类脑计算

"}, bioImg=KhPhUJkZgXzeQ+KozAYuwQ==, bioContent=

1995年生。博士。主要研究方向为机器视觉、类脑计算

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1329420413156831496, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, xref=1, ext=[AuthorCompanyExt(id=1329420413165220105, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, companyId=1329420413156831496, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1Qiyuan Lab, Beijing 100095), AuthorCompanyExt(id=1329420413173608714, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, companyId=1329420413156831496, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1启元实验室 北京 100095)])]), Author(id=1329420413655953690, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, orderNo=1, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=null, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1329420413739839772, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, authorId=1329420413655953690, language=EN, stringName=Jingyue Zhao, firstName=Jingyue, middleName=null, lastName=Zhao, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=2, address=2Institute of National Defense Science and Technology Innovation, Beijing 100091, bio={"content":"

born in 1994. PhD, assistant professor. Member of CCF. Her main research interests include neuromorphic computing, brain-inspired robotics, and embodied intelligence

"}, bioImg=null, bioContent=

born in 1994. PhD, assistant professor. Member of CCF. Her main research interests include neuromorphic computing, brain-inspired robotics, and embodied intelligence

, aboutCorrespAuthor=null), CN=AuthorExt(id=1329420413823725853, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, authorId=1329420413655953690, language=CN, stringName=赵静月, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=2, address=2军事科学院国防科技创新研究院 北京 100091, bio={"img":"yFKDKr6xdq4JOe65N7A/7w==","content":"

1994年生。博士,助理研究员。CCF会员。主要研究方向为类脑计算、类脑机器人、具身智能

"}, bioImg=yFKDKr6xdq4JOe65N7A/7w==, bioContent=

1994年生。博士,助理研究员。CCF会员。主要研究方向为类脑计算、类脑机器人、具身智能

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1329420413232328971, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, xref=2, ext=[AuthorCompanyExt(id=1329420413236523276, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, companyId=1329420413232328971, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2Institute of National Defense Science and Technology Innovation, Beijing 100091), AuthorCompanyExt(id=1329420413244911885, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, companyId=1329420413232328971, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2军事科学院国防科技创新研究院 北京 100091)])]), Author(id=1329420413899223327, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, orderNo=2, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=null, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1329420413966332193, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, authorId=1329420413899223327, language=EN, stringName=Ge Zhang, firstName=Ge, middleName=null, lastName=Zhang, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=3, address=3Beijing Institute of Tracking and Telecommunication Technology, Beijing 100094, bio={"content":"

born in 1996. PhD, assistant professor. Her main research interests include hyperspectral image interpretation and multimodal image fusion

"}, bioImg=null, bioContent=

born in 1996. PhD, assistant professor. Her main research interests include hyperspectral image interpretation and multimodal image fusion

, aboutCorrespAuthor=null), CN=AuthorExt(id=1329420414062801186, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, authorId=1329420413899223327, language=CN, stringName=张歌, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=3, address=3北京跟踪与通信技术研究所 北京 100094, bio={"img":"ktZjwGCCMGbJEGqOAwMC+g==","content":"

1996年生。博士,助理研究员。主要研究方向为高光谱图像解译、多模态图像融合

"}, bioImg=ktZjwGCCMGbJEGqOAwMC+g==, bioContent=

1996年生。博士,助理研究员。主要研究方向为高光谱图像解译、多模态图像融合

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1329420413303632142, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, xref=3, ext=[AuthorCompanyExt(id=1329420413307826447, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, companyId=1329420413303632142, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=3Beijing Institute of Tracking and Telecommunication Technology, Beijing 100094), AuthorCompanyExt(id=1329420413316215056, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, companyId=1329420413303632142, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=3北京跟踪与通信技术研究所 北京 100094)])]), Author(id=1329420414188630309, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, orderNo=3, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=null, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1329420414268322087, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, authorId=1329420414188630309, language=EN, stringName=Bin Fan, firstName=Bin, middleName=null, lastName=Fan, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=4, address=4Shaanxi Key Laboratory of Information Acquisition and Processing (Northwestern Polytechnical University), Xi’an 710129, bio={"content":"

born in 1994. PhD, professor, PhD supervisor. His main research interests include machine vision, computational photography, and 3D reconstruction

"}, bioImg=null, bioContent=

born in 1994. PhD, professor, PhD supervisor. His main research interests include machine vision, computational photography, and 3D reconstruction

, aboutCorrespAuthor=null), CN=AuthorExt(id=1329420414335430952, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, authorId=1329420414188630309, language=CN, stringName=樊斌, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=4, address=4陕西省信息获取与处理重点实验室(西北工业大学) 西安 710129, bio={"img":"0CT0S8s/+xpmgX1cS2GoNQ==","content":"

1994年生。博士,教授,博士生导师。主要研究方向为机器视觉、计算摄影、3维重建

"}, bioImg=0CT0S8s/+xpmgX1cS2GoNQ==, bioContent=

1994年生。博士,教授,博士生导师。主要研究方向为机器视觉、计算摄影、3维重建

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1329420413366546705, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, xref=4, ext=[AuthorCompanyExt(id=1329420413379129618, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, companyId=1329420413366546705, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=4Shaanxi Key Laboratory of Information Acquisition and Processing (Northwestern Polytechnical University), Xi’an 710129), AuthorCompanyExt(id=1329420413383323923, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, companyId=1329420413366546705, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=4陕西省信息获取与处理重点实验室(西北工业大学) 西安 710129)])]), Author(id=1329420414444482858, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, orderNo=4, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=null, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1329420414528368940, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, authorId=1329420414444482858, language=EN, stringName=Zhijie Yang, firstName=Zhijie, middleName=null, lastName=Yang, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=2, address=2Institute of National Defense Science and Technology Innovation, Beijing 100091, bio={"content":"

born in 1995. PhD. His main research interests include neuromorphic computing and computer architecture

"}, bioImg=null, bioContent=

born in 1995. PhD. His main research interests include neuromorphic computing and computer architecture

, aboutCorrespAuthor=null), CN=AuthorExt(id=1329420414595477805, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, authorId=1329420414444482858, language=CN, stringName=杨智杰, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=2, address=2军事科学院国防科技创新研究院 北京 100091, bio={"img":"7ewoPcccusAQ2CwmLo8OpA==","content":"

1995 年生。博士。主要研究方向为类脑计算、计算机体系结构

"}, bioImg=7ewoPcccusAQ2CwmLo8OpA==, bioContent=

1995 年生。博士。主要研究方向为类脑计算、计算机体系结构

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1329420413232328971, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, xref=2, ext=[AuthorCompanyExt(id=1329420413236523276, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, companyId=1329420413232328971, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2Institute of National Defense Science and Technology Innovation, Beijing 100091), AuthorCompanyExt(id=1329420413244911885, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, companyId=1329420413232328971, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2军事科学院国防科技创新研究院 北京 100091)])]), Author(id=1329420414658392367, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, orderNo=5, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=null, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1329420414733889841, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, authorId=1329420414658392367, language=EN, stringName=Zhong Wan, firstName=Zhong, middleName=null, lastName=Wan, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=2, address=2Institute of National Defense Science and Technology Innovation, Beijing 100091, bio={"content":"

born in 1989. PhD, associate professor. Her main research interest includes neuromorphic computing

"}, bioImg=null, bioContent=

born in 1989. PhD, associate professor. Her main research interest includes neuromorphic computing

, aboutCorrespAuthor=null), CN=AuthorExt(id=1329420414805193010, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, authorId=1329420414658392367, language=CN, stringName=万众, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=2, address=2军事科学院国防科技创新研究院 北京 100091, bio={"img":"1fyh1oBMlUaSmeI3QF5z7g==","content":"

1989年生。博士,副研究员。主要研究方向为类脑计算

"}, bioImg=1fyh1oBMlUaSmeI3QF5z7g==, bioContent=

1989年生。博士,副研究员。主要研究方向为类脑计算

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1329420413232328971, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, xref=2, ext=[AuthorCompanyExt(id=1329420413236523276, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, companyId=1329420413232328971, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2Institute of National Defense Science and Technology Innovation, Beijing 100091), AuthorCompanyExt(id=1329420413244911885, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, companyId=1329420413232328971, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2军事科学院国防科技创新研究院 北京 100091)])]), Author(id=1329420414868107572, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, orderNo=6, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=null, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1329420414939410742, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, authorId=1329420414868107572, language=EN, stringName=Yuchao Dai, firstName=Yuchao, middleName=null, lastName=Dai, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=4, address=4Shaanxi Key Laboratory of Information Acquisition and Processing (Northwestern Polytechnical University), Xi’an 710129, bio={"content":"

born in 1982. PhD, professor, PhD supervisor. Senior member of CCF. His main research interests include machine vision and embodied intelligence

"}, bioImg=null, bioContent=

born in 1982. PhD, professor, PhD supervisor. Senior member of CCF. His main research interests include machine vision and embodied intelligence

, aboutCorrespAuthor=null), CN=AuthorExt(id=1329420415010713911, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, authorId=1329420414868107572, language=CN, stringName=戴玉超, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=4, address=4陕西省信息获取与处理重点实验室(西北工业大学) 西安 710129, bio={"img":"+IQ/KZFMXiIksA2+vA8IPg==","content":"

1982年生。博士,教授,博士生导师。CCF高级会员。主要研究方向为机器视觉、具身智能

"}, bioImg=+IQ/KZFMXiIksA2+vA8IPg==, bioContent=

1982年生。博士,教授,博士生导师。CCF高级会员。主要研究方向为机器视觉、具身智能

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1329420413366546705, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, xref=4, ext=[AuthorCompanyExt(id=1329420413379129618, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, companyId=1329420413366546705, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=4Shaanxi Key Laboratory of Information Acquisition and Processing (Northwestern Polytechnical University), Xi’an 710129), AuthorCompanyExt(id=1329420413383323923, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, companyId=1329420413366546705, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=4陕西省信息获取与处理重点实验室(西北工业大学) 西安 710129)])]), Author(id=1329420415090405689, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, orderNo=7, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=leiwang@nudt.edu.cn, emailSecond=null, emailThird=null, correspondingAuthor=1, authorType=1, ext={EN=AuthorExt(id=1329420415170097467, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, authorId=1329420415090405689, language=EN, stringName=Lei Wang, firstName=Lei, middleName=null, lastName=Wang, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=2, *, address=2Institute of National Defense Science and Technology Innovation, Beijing 100091, bio={"content":"

born in 1977. PhD, professor, PhD supervisor. Distinguished member of CCF. Her main research interests include computer architecture and intelligent chips

"}, bioImg=null, bioContent=

born in 1977. PhD, professor, PhD supervisor. Distinguished member of CCF. Her main research interests include computer architecture and intelligent chips

, aboutCorrespAuthor=null), CN=AuthorExt(id=1329420415233012028, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, authorId=1329420415090405689, language=CN, stringName=王蕾, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=2, *, address=2军事科学院国防科技创新研究院 北京 100091, bio={"img":"npIuc8x7vKLsZcr8zjStkg==","content":"

1977年生。博士,研究员,博士生导师。CCF杰出会员。主要研究方向为计算机体系结构、智能芯片

"}, bioImg=npIuc8x7vKLsZcr8zjStkg==, bioContent=

1977年生。博士,研究员,博士生导师。CCF杰出会员。主要研究方向为计算机体系结构、智能芯片

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1329420413232328971, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, xref=2, ext=[AuthorCompanyExt(id=1329420413236523276, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, companyId=1329420413232328971, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2Institute of National Defense Science and Technology Innovation, Beijing 100091), AuthorCompanyExt(id=1329420413244911885, tenantId=1146029695717560320, journalId=1324759914146529478, articleId=1329420410619277551, companyId=1329420413232328971, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2军事科学院国防科技创新研究院 北京 100091)])])]
王宇飞1 , 赵静月2, 张歌3, 樊斌4, 杨智杰2, 万众2, 戴玉超4, 王蕾2, *
作者信息
  • 1启元实验室 北京 100095
  • 2军事科学院国防科技创新研究院 北京 100091
  • 3北京跟踪与通信技术研究所 北京 100094
  • 4陕西省信息获取与处理重点实验室(西北工业大学) 西安 710129
通讯作者:
作者简介:

作者贡献声明

王宇飞负责研究方案实施、实验开展与结果分析,并完成论文撰写;赵静月参与算法实验验证,负责实验结果整理及论文内容修订;张歌负责研究思路把关、理论分析指导及论文审阅;樊斌参与理论框架论证,并对论文内容和技术表述进行修改完善;杨智杰负责实验方案设计、实验流程优化及相关验证工作;万众负责论文审校、内容修改及文字规范化;戴玉超为研究提供理论指导,并参与论文结构调整与学术内容修订;王蕾负责相关理论问题指导,并对论文内容进行审阅和完善。

1995年生。博士。主要研究方向为机器视觉、类脑计算

1994年生。博士,助理研究员。CCF会员。主要研究方向为类脑计算、类脑机器人、具身智能

1996年生。博士,助理研究员。主要研究方向为高光谱图像解译、多模态图像融合

1994年生。博士,教授,博士生导师。主要研究方向为机器视觉、计算摄影、3维重建

1995 年生。博士。主要研究方向为类脑计算、计算机体系结构

1989年生。博士,副研究员。主要研究方向为类脑计算

1982年生。博士,教授,博士生导师。CCF高级会员。主要研究方向为机器视觉、具身智能

1977年生。博士,研究员,博士生导师。CCF杰出会员。主要研究方向为计算机体系结构、智能芯片

SpikingReLU: A High-Performance Spiking Neural Network Based on Hybrid Model Conversion
Yufei Wang1 , Jingyue Zhao2, Ge Zhang3, Bin Fan4, Zhijie Yang2, Zhong Wan2, Yuchao Dai4, Lei Wang2, *
Affiliations
  • 1Qiyuan Lab, Beijing 100095
  • 2Institute of National Defense Science and Technology Innovation, Beijing 100091
  • 3Beijing Institute of Tracking and Telecommunication Technology, Beijing 100094
  • 4Shaanxi Key Laboratory of Information Acquisition and Processing (Northwestern Polytechnical University), Xi’an 710129
  • born in 1995. PhD. His main research interests include machine vision and neuromorphic computing

    born in 1994. PhD, assistant professor. Member of CCF. Her main research interests include neuromorphic computing, brain-inspired robotics, and embodied intelligence

    born in 1996. PhD, assistant professor. Her main research interests include hyperspectral image interpretation and multimodal image fusion

    born in 1994. PhD, professor, PhD supervisor. His main research interests include machine vision, computational photography, and 3D reconstruction

    born in 1995. PhD. His main research interests include neuromorphic computing and computer architecture

    born in 1989. PhD, associate professor. Her main research interest includes neuromorphic computing

    born in 1982. PhD, professor, PhD supervisor. Senior member of CCF. His main research interests include machine vision and embodied intelligence

    born in 1977. PhD, professor, PhD supervisor. Distinguished member of CCF. Her main research interests include computer architecture and intelligent chips

出版时间: 2026-08-01 doi: 10.7544/issn1000-1239.202550722
文章导航
收藏切换

受大脑启发的脉冲神经网络(spiking neural networks,SNNs)由于其强大的时空编码能力和事件驱动计算等特性展现出巨大的发展潜力。目前,主流的SNNs方法分为将训练完备的人工神经网络(artificial neural networks,ANNs)转换为SNNs以及直接训练SNNs这2类。然而,现有的ANNs转换方法需要较长的时间步减少转换误差,直接训练的SNNs则受限于网络表征能力的不足。针对上述挑战,提出一种基于混合模型转换的高性能脉冲神经网络。所提方法将模型训练与推理解耦:在训练阶段,策略性地将SNNs中部分脉冲神经元替换为ReLU激活函数,增强网络的特征学习能力,从而构建一个同时包含ANNs与SNNs的高性能混合模型;由于混合模型中的ANNs部分无法进行事件驱动计算,进一步引入重参数化技巧,在不损失性能的前提下,将混合模型的推理过程全部转换为事件驱动计算,从而使所提方法兼具高性能和事件驱动计算的优势。实验结果表明:所提方法在CIFAR-10和CIFAR-100数据集上的分类准确率分别为97.31%和83.34%,在ImageNet数据集上的分类准确率为70.89%,在神经静态数据集CIFAR10-DVS的分类准确率为82.71%,均优于现有国内外最先进的方法。

类脑计算  /  脉冲神经网络  /  事件驱动计算  /  混合模型  /  训练-推理解耦方法

Brain-inspired spiking neural networks (SNNs) have shown considerable promise due to their strong spatiotemporal encoding capabilities and event-driven computation. Current mainstream SNN approaches can be divided into two categories: converting pre-trained artificial neural networks (ANNs) into SNNs (ANN conversion methods) and directly training SNNs. However, ANN conversion methods require large time steps to reduce conversion errors, while directly trained SNNs suffer from poor representational capacity. To address these challenges, we propose a high-performance spiking neural network based on hybrid model conversion. The proposed method decouples model training and inference. During the training phase, a portion of the spiking neurons in the SNNs are strategically replaced with ReLU activation functions to enhance the feature learning ability, thereby constructing a high-performance hybrid model that integrates both ANN and SNN components. Since the ANN part of the hybrid model cannot perform event-driven computation, we further introduce a reparameterization technique to convert the entire inference process of the hybrid model into event-driven computation without performance loss. Therefore, the proposed method combines the advantages of high performance and event-driven computation. Experimental results show that the proposed method achieves classification accuracies of 97.31% and 83.34% on CIFAR-10 and CIFAR-100, respectively, 70.89% on ImageNet, and 82.71% on the neuromorphic dataset CIFAR10-DVS, outperforming state-of-the-art methods.

brain-inspired computing  /  spiking neural networks  /  event-driven computing  /  hybrid model  /  training-inference decoupling method
王宇飞, 赵静月, 张歌, 樊斌, 杨智杰, 万众, 戴玉超, 王蕾. SpikingReLU:一种基于混合模型转换的高性能脉冲神经网络. 计算机研究与发展, 2026 , 63 (8) : 2143 -2153 . DOI: 10.7544/issn1000-1239.202550722
Yufei Wang, Jingyue Zhao, Ge Zhang, Bin Fan, Zhijie Yang, Zhong Wan, Yuchao Dai, Lei Wang. SpikingReLU: A High-Performance Spiking Neural Network Based on Hybrid Model Conversion[J]. Journal of Computer Research and Development, 2026 , 63 (8) : 2143 -2153 . DOI: 10.7544/issn1000-1239.202550722
类脑计算旨在借鉴生物神经系统的结构特征与信息处理机制,突破传统冯·诺依曼架构在能效方面的瓶颈,构建新一代高效智能计算体系[1]。脉冲神经网络(spiking neural networks,SNNs)凭借其高度仿生的神经元模型和事件驱动计算机制,受到国内外研究者的广泛关注,是类脑计算领域的重要研究方向[2]。相对于人工神经网络(artificial neural networks,ANNs)中的非线性激活层[3],SNNs使用脉冲神经元作为激活层,赋予了SNNs强大的时空编码能力和事件驱动计算等特性。
主流的SNNs研究大致分为ANNs转换方法与直接训练SNNs这2类。然而,现有的ANNs转换方法依赖数百个时间步来减小转换误差,导致推理效率低下;直接训练的SNNs受限于现有神经元较弱的表征能力,难以突破性能瓶颈。为提升SNNs的性能,国内外研究学者提出学习神经元的关键参数,如脉冲阈值[4]和膜时间常数[5]。Ternary spike[6]和I-LIF[7]提出将2值神经元改进为3值和整数值。受ANNs启发,基于SNNs的残差学习网络[8-9]和注意力机制[10]也相继被提出以缓解网络深度退化问题并提高性能。此外,一些研究者提出施加人为设计的正则化[11]或使用性能较强的ANNs模型进行知识蒸馏[12]。
尽管现有研究已取得显著进展,但存在一个关键问题尚未得到充分重视。如图1(a)所示,主流的SNNs网络主干通常叠加多个“卷积-批量归一化-脉冲神经元”(convolution batch-normal spiking-neuron,CBS)模块,然而,本文实验发现,多个堆叠的CBS模块并不能有效提高SNNs性能。具体而言,CBS模块中的卷积层应尽可能从脉冲输入中提取高级特征。但该模块中的单层卷积却只能执行简单的线性操作,学习能力有限,无法有效地从脉冲输入中提取复杂特征。同时,CBS模块的神经元又将卷积层的输出重新量化为脉冲特征,导致后续的CBS模块会遇到相同的限制。
为解决上述问题,本文提出一种基于混合模型转换的脉冲神经网络——SpikingReLU,使SNNs具备高性能和事件驱动计算的双重优势。具体而言,SpikingReLU策略性地将多个连续CBS模块序列中除最后一个神经元之外的所有神经元替换为ReLU激活层,其中最后一个神经元确保模块的输出仍为脉冲信号。多个结合非线性ReLU激活层的卷积层赋予了网络强大的特征提取能力,但也破坏SNNs事件驱动计算的优势。如图1(b)所示,针对该问题,本文进一步提出一种训练-推理解耦策略,首先训练得到一个同时包含ANNs和SNNs的高性能混合网络模型;在推理阶段,采用重参数化技巧重构模型计算过程,将其完全转换为事件驱动计算。
综上所述,本文的主要贡献如下:
1)提出了一个同时包含ANNs和SNNs的混合模型,策略性地将现有SNNs中的部分神经元替换为非线性ReLU激活层,显著提高了网络的特征表征能力;
2)提出了一种无损的混合模型转换方法,在模型推理阶段将混合模型完全转换为事件驱动计算,从而使混合模型兼具高性能和事件驱动计算的优势;
3)在多个主流的静态数据集与神经形态数据集上的实验结果表明,本文方法的性能优于现有国内外最先进方法。
ANNs转换方法通过将训练完备的ANNs转换为SNNs,从而快速且有效地获取SNNs模型。该领域开创性的工作[13]验证了ReLU激活函数与脉冲神经元发放率之间的等价性,并将学习到的ANNs模型权重转移到SNNs模型中。文献[14-15]首先移除ANNs网络的局部响应归一化层,将最大值池化改为平均池化,之后训练使用软化泄漏积分发放(leaky integrate-and-fire,LIF)响应函数作为非线性层的网络,并在训练过程中添加噪声增强网络的鲁棒性,最后使用LIF脉冲模型替换软化的LIF率模型,从而将ANNs网络转换为SNNs。之后,文献[16]提出在ANNs训练阶段使用零偏差的修正线性单元,在推理阶段使用一种新的权重归一化方法来帮助调节神经元发放率,从而减少卷积网络和全连接深度网络在转换过程中的性能损失。为进一步减少转换误差,SATP[17]提出了一种自驱动自适应阈值可塑性机制,其中神经元使用无监督学习规则并根据其自身的状态信息自主调整发放阈值。为减少转换之后SNNs的时间步长,文献[18]对ANNs到SNNs的转换过程进行了理论分析,表明缩放阈值发挥着与权重正则化类似的作用,基于此理论结果,提出了优化初始膜电位减少每层的转换损失,将时间步缩小为32。
由于SNNs中的神经元使用阶跃函数判定是否发射脉冲,阶跃函数在脉冲发射点处的导数为无穷大,而其他地方的导数为0,这导致SNNs无法直接使用反向传播对权重进行更新。针对该问题,现有方法在反向传播时使用一个光滑可微的替代函数来近似不可微的阶跃函数,使SNNs可以直接训练。直接训练的SNNs模型主要从神经元、网络结构和训练技术3个层面提升脉冲神经网络的性能[2]。在神经元层面,研究人员提出了PLIF[5],GLIF[19]等多种改进神经元模型。在网络结构层面,受ANNs中有效网络结构的启发,SNNs领域出现了脉冲版本的残差学习网络[8-9]、注意力机制[10]和Transformer架构[20],显著提升了网络的性能。在训练技术层面,研究者通过引入辅助损失函数作为正则化项[11],或在学习阶段使用更强大的ANNs教师模型来引导SNNs模型进行知识蒸馏[12]。此外,Real spike[21]和ILIF[7]等方法提出了训练-推理解耦的策略,在保持事件驱动优势的同时,有效提升了直接训练SNNs的性能。
神经元是脉冲神经网络的核心计算单元,其主要功能是整合时空特征并将其量化为脉冲。其中,LIF神经元因其可以较好地平衡生物合理性和计算复杂度,是目前最常用的神经元模型。如式(1)所示,LIF神经元的神经学动力过程包含3个步骤:充电(charging)、发放(firing)和重置(resetting)。具体而言,在时间步$ t $,来自前一时刻的膜电位V[t−1]与当前的空间输入信息X[t]被整合到当前膜电位H[t]。脉冲输出B[t]由阶跃函数Θ(·)决定:当膜电位H[t]超过发放阈值$ {\boldsymbol V}_{\text{threshold}} $时,LIF神经元产生一个脉冲(B[t] = 1),否则保持静息状态(B[t] = 0)。发放脉冲会消耗神经元累积的电荷,导致膜电位瞬时下降,即膜电位重置过程。以常用的软重置为例,若神经元发放脉冲,则从膜电位中减去阈值电压$ {\boldsymbol V}_{\rm threshold} $。为简化表达,本文将LIF神经元的充电过程记作函数$Charging(\cdot) $,发放过程记作函数Firing (·),重置过程记作函数$Resetting(\cdot) $。
$ LIF=\left\{\begin{aligned}&Charging\text{:}{\boldsymbol{H}}[t]={\boldsymbol{V}}[t-1]+{\boldsymbol{X}}[t]\;,\\&Firing\text{:}{\boldsymbol{B}}[t]=\mathit{\Theta }\left({\boldsymbol{H}}[t]-{\boldsymbol V}_{\rm threshold}\right)\;,\\&Resetting\text{:}\boldsymbol V[t]=\boldsymbol H[t]-{\boldsymbol V}_{\rm threshold}{\boldsymbol{B}}[t]\;。\end{aligned}\right. $
基于目前成熟的ANNs,SNNs致力于用LIF神经元替代ANNs中的非线性激活层。因此,与ANNs中常见的“卷积-批归一化-非线性激活”(convolution batch-normal non-linear,CBN)特征提取模块不同,SNNs通常由多个堆叠的CBS模块构成。
对于一个CBS特征提取模块,假设输入特征图为B0,其特征图的高度和宽度分别为H和W,通道数为C。CBS模块中的卷积层的核为K1,卷积核的维度为KKCC,其中K为卷积核尺寸大小,输入与输出特征图的通道数都为C,该卷积层的输出特征图F1中每个元素的计算方式为:
$ \boldsymbol{F}_{k,l,n}^{1}=\sum \limits_{i,j,m}\boldsymbol{K}_{i,j,m,n}^{1}\cdot \boldsymbol{B}_{k+i-1,l+j-1,m}^{0}\;, $
其中输入特征图B0通常是由前一个CBS模块生成或为输入信号经过脉冲编码得到的二进制脉冲特征。在神经形态硬件上,CBS模块中的卷积运算可以通过累加(accumulation,AC)操作实现,相比于传统卷积的乘加(multiply-accumulation,MAC)操作更加节能[22-24],这是SNNs实现高能效的一个关键因素。为了确保训练稳定性,卷积层之后会使用一个批归一化(batch normalization,BN)层。但由于BN层在推理过程中可以与前一层卷积融合为一个带偏置的卷积层,本文不对BN层进行单独描述。最后,卷积层的输出被送入LIF神经元,将浮点特征重新量化为二进制脉冲特征,其过程如式(3)所示。
$ \boldsymbol{B}_{}^{1}=Firing(\boldsymbol{F}^{1})\;。 $
传统的CBS脉冲特征提取模块存在2个缺陷:首先,卷积层的输出特征量化为脉冲信号的过程会产生显著的量化误差,阻碍特征信息的传播;其次,单层卷积是一个学习能力有限的线性操作,无法有效地从脉冲输入信号中提取到高级特征。为解决上述问题,本文提出一种混合ANNs和SNNs特性的脉冲特征提取模块—SpikingReLU,提出的模型可作为现有CBS脉冲特征提取模块的改进。以改进2个连续的CBS模块为例,SpikingReLU将CBS模块序列中除最后一个神经元之外的所有神经元替换为非线性ReLU激活层,其中最后一个神经元确保SpikingReLU模块的输出仍为脉冲信号。如式(4)所示,SpikingReLU将第1个CBS模型中卷积层的输出特征F1输入到ReLU激活层中,得到激活值G1。
$ {\boldsymbol{G}}^{1}=ReLU ({\boldsymbol{F}}^{1} )\;。 $
之后,将其输入到第2个CBS模块的卷积层中,假定该卷积层的核为K2,输出特征记为F2,计算过程如式(5)所示。
$ \boldsymbol{F}_{k,l,n}^{2}=\sum \limits_{i,j,m}\boldsymbol{K}_{i,j,m,n}^{2}\cdot \boldsymbol{G}_{k+i-1,l+j-1,m}^{1}\;, $
其中ReLU激活层的输出特征G1由浮点数组成,因此后续的CBS模块中的卷积层均无法以事件驱动的方式计算。
为解决上述问题,本文提出一种训练-推理解耦策略,如图2所示。在模型训练阶段,SpikingReLU按照式(2)(4)(5)顺序执行。在模型推理阶段,SpikingReLU采用重参数化技术将混合的ANNs和SNN模型转换为完全事件驱动的计算方式,从而使模型兼具高性能和事件驱动计算的优势。
具体而言,在模型推理阶段,首先改变非线性ReLU激活层原本的计算方式。对于输入特征F1,ReLU激活层的功能是将其所有负值转换为零,同时保持正值不变。如式(6)所示,F1的正值掩码M1可以通过将发射阈值设为0的发射操作获得。
$ {\boldsymbol{M}}^{1}=Firin{g}_{\rm threshold=0} ({\boldsymbol{F}}^{1} )\;。 $
因此,与ReLU激活层原本的计算方式不同,其激活特征G1可以通过式(7)计算得到。
$ {\boldsymbol{G}}^{1}={\boldsymbol{F}}^{1}\odot {\boldsymbol{M}}^{1}\;, $
其中$ \odot $操作表示逐元素相乘。
为了在模型推理阶段,将SpikingReLU转换为完全的事件驱动计算,首先使用式(7)中改进后的ReLU层计算方法替换其原本计算方法,然后将式(7)代入式(5)中,得到第2个CBS模块卷积层的输出特征F2,具体计算过程如式(8)所示。
$ \boldsymbol{F}_{k,l,n}^{2}=\sum \limits_{i,j,m}\boldsymbol{K}_{i,j,m,n}^{2}\cdot \left(\boldsymbol{F}_{k+i-1,l+j-1,m}^{1}\cdot \boldsymbol{M}_{k+i-1,l+j-1,m}^{1}\right)\;。 $
根据乘法交换律,交换式(8)中乘法的顺序,得到式(9)。
$ \boldsymbol{F}_{k,l,n}^{2}=\sum \limits_{i,j,m}\left(\boldsymbol{K}_{i,j,m,n}^{2}\cdot \boldsymbol{F}_{k+i-1,l+j-1,m}^{1}\right)\cdot \boldsymbol{M}_{k+i-1,l+j-1,m}^{1}\;, $
其中$ \boldsymbol{F}_{k+i-1,l+j-1,m}^{1} $是通过式(2)计算得到。因此,将式(2)代入式(9)中可得式(10)。
$ \begin{split}&\boldsymbol{F}_{k,l,n}^{2}=\sum \limits_{i,j,m}\left(\boldsymbol{K}_{i,j,m,n}^{2} \left(\sum \limits_{{i}^{\prime},{j}^{\prime},{m}^{\prime}}\boldsymbol{K}_{{i}^{\prime},{j}^{\prime},{m}^{\prime},{n}^{\prime}}^{1}\boldsymbol{B}_{k+i-1+{i}^{\prime}-1,l+j-1+{j}^{\prime}-1,{m}^{\prime}}^{0}\right) \right) \times \\&\boldsymbol{M}_{k+i-1,l+j-1,m}^{1}\;。\end{split}$
在式(10)中,脉冲输入信号B0连续与卷积核K1和卷积核K2进行卷积运算,这2个连续的卷积层可以通过合并重构为1个卷积核$ \widetilde{\boldsymbol{K}} $,因此计算过程可整理为如下所示:
$ \boldsymbol{F}_{k,l,n}^{2}=\sum \limits_{i,j,m,{i}^{\prime},{j}^{\prime},{m}^{\prime}}\left({\widetilde{\boldsymbol{K}}}_{{{i}^{\prime}},{{j}^{\prime}},{{m}^{\prime}},n}\cdot \boldsymbol{B}_{k+{i}^{\prime}-1,l+{j}^{\prime}-1,{m}^{\prime}}^{0}\right)\cdot \boldsymbol{M}_{k+i-1,l+j-1,m}^{1}\;。 $
在式(11)中,特征B0和M1都由二进制脉冲信号组成,因此,SpikingReLU在推理阶段,所有运算都可以采用事件驱动的计算方式执行。
3.1节和3.2节以2个连续的CBS脉冲特征提取模块为例,介绍了SpikingReLU改进方法。值得一提的是,SpikingReLU对于多个连续的CBS模块同样有效。具体而言,在模型训练阶段,SpikingReLU将CBS模块序列中除最后一个之外的神经元替换为ReLU激活层,并在推理阶段采用重参数化技术将SpikingReLU转换为纯事件驱动计算。表1的实验结果表明:当模块数量增加时,SpikingReLU模块比传统CBS模块获得了更显著的性能提升,能够更有效地缓解网络深度退化问题。此外,实验结果还表明SpikingReLU模块中还可采用一些改进的LIF神经元,如3值脉冲神经元和I-LIF神经元,协同提升SNNs的性能。
本文提出的SpikingReLU模块作为对传统CBS模块的改进,可用于替换现有SNNs网络中2个或多个连续的CBS模块。例如,SpikingReLU模块可在输入信号脉冲编码阶段使用,也可用于基于SNNs的残差网络中学习残差特征图。
本节对比了提出的SpikingReLU与现有的相关工作。早期研究[14-15]提出一种替换LIF神经元的训练方法,首先修改ANNs的网络结构,使用软化LIF神经元模型并配合噪声训练得到静态网络,之后将软化LIF神经元模型替换为真正的LIF神经元,得到性能相近的SNNs。与之相比,SpikingReLU不需要对ANNs网络进行修改,在训练阶段策略性地使用ReLU替换LIF神经元,在推理阶段使用重参数技巧实现脉冲驱动计算。Real spike[21]首次提出了一种训练-推理解耦的脉冲神经网络。在训练阶段,Real spike为2值脉冲预测一个系数图以获得实数值脉冲;在推理阶段,该系数图可以通过使用重参数化技术被吸收到后续的卷积核中。Real spike中的系数图在训练阶段中通过学习优化,在推理阶段保持不变。相比于提出的SpikingReLU,特征图F1工作原理类似于Real spike的系数图。但是,F1不同于Real spike的可学习的超参数,而是由前一层卷积预测得到。因此,SpikingReLU的系数图F1可根据输入样本进行自适应调整,从而进一步增强SNNs的表征能力。
为了全面验证SpikingReLU的优越性,本文在RGB图像和事件流2种不同数据进行实验。在RGB图像分类实验中,选择了当前主流的图像分类数据集,其中包括CIFAR-10和CIFAR-100以及ImageNet数据集。在对事件流实验中,选择了常用的CIFAR10-DVS数据集。遵循通用的评估方法,将计算分类准确率作为分类方法的主要性能指标。
本文模型使用Pytorch实现,并在GeForce RTX 3090 GPU上进行实验。本文的所有实验均使用SGD优化器,动量设置为0.9,学习率遵循余弦退火衰减策略。对于RGB图像分类任务,批大小(batch size)设置为128,SpikingReLU在CIFAR-10和CIFAR-100数据集上训练400轮,在ImageNet数据集上训练100轮。此外,本文还采用一些常见策略来提高方法的性能,包括自动数据增强和测试时增强。对于事件数据,使用SpikingJelly软件库加载事件数据,批大小设置为12,在CIFAR10-DVS数据集上训练200轮。
表1中对比了基于SNNs的残差学习网络模块以及传统CBS模块的性能。实验结果表明,当残差学习模块仅使用单个CBS模块学习残差图时,分类准确率为95.82%,当将CBS模块数量增加到2时,准确率为96.19%,性能提升相对较小。表1同时表明,由于实数值LIF神经元比2值LIF具有更强的特征表达能力,使用Real spike可以提高方法的准确率。值得注意地是,使用本文所提出的SpikingReLU改进2个连续的CBS模块,SNNs的性能会显著增强,这是由于SpikingReLU采用多个卷积层和非线性层可以有效地从脉冲输入中学习高级特征表示。改进之后方法的准确率为97.05%,比使用CBS模块的方法高0.86个百分点,比使用CBS与Real spike结合的方法高0.75个百分点。同时,随着模块数量的增加,SpikingReLU的性能提升幅度更加显著,进一步证明了其有效性。此外,图3中可视化了CBS模块和SpikingReLU模块在不同时间步的中间特征的同一个通道特征。从中观察到,CBS模块的特征是离散的,仅由0和1组成,从而限制了CBS模块之间信息的传递,对比而言,SpikingReLU模块中的特征更加丰富。
对于SNNs,更大的时间步意味着更强的时空特征提取能力。遵循现有工作,本文将时间步设置为1,2,4,在CIFAR-10和CIFAR-100数据集上的实验结果如表2所示。表2表明,随着时间步的增加,Spiking-ReLU的准确率持续提高。值得注意的是,Spiking-ReLU在相同时间步下的性能均优于现有国内外最先进的方法,这充分验证了SpikingReLU在相同时间步设置下能够更好地利用时空特征。
近年来,国内外的研究学者提出了一系列改进的LIF神经元来提升SNNs的性能,其中最具代表性的有3值脉冲神经元和I-LIF神经元等。为验证SpikingReLU与改进的LIF神经元的兼容性,本节验证了SpikingReLU中采用3值和4值的I-LIF神经元在CIFAR-10数据集的性能。表1中的实验结果表明,当神经元从2值变为3值时,准确率从97.31%提高到97.47%,表明SpikingReLU可以与现有的改进脉冲神经元协同作用,共同提升SNNs的网络表达能力。当SpikingReLU采用4值神经元时,方法的性能提升有限,这表明改进神经元带来的提升正在接近饱和。
为验证SpikingReLU在性能上的优越性,本文在CIFAR-10,CIFAR-100,ImageNet这3个主流的图像分类数据集上,对比SpikingReLU与国内外先进方法的性能。本文设计的图像分类网络采用ResNet网络架构,并在残差学习分支中使用SpikingReLU模块。在CIFAR-10和CIFAR-100数据集上,SpikingReLU的模型大小与ResNet-18相当,在ImageNet数据集上,SpikingReLU的模型大小与ResNet-34相当。本文在CIFAR-10和CIFAR-100数据集上将时间步设置为1,2,4,在ImageNet数据集,时间步设置为4。为了表述简洁,本文只报告了对比方法的最优结果:表2表明,SpikingReLU在CIFAR-10数据集上达到了97.31%的准确率,超越了所有对比方法,在CIFAR-100数据集上达到了83.34%的最高准确率,同样排名第1。值得注意地是,SpikingReLU使用所有时间步的准确率均优于现有国内外最优的方法。同时,表3对比了TET[35],RecDis-SNN[36],GLIF[19],Real spike[21],RMP-loss[37],MPBN[38],Ternary Spike[6],AGMM[33],STAA-SNN[34],FATA-SNN[39]算法在ImageNet数据集的实验结果。表3表明,SpikingReLU在ImageNet数据集上也达到了70.89%的最高准确率,进一步证明了SpikingReLU的有效性。
构建可以部署在神经形态硬件上处理来自动态视觉传感器(dynamic vision sensor, DVS)数据的脉冲神经网络是一个具有巨大应用潜力的研究方向。因此,本文进一步地在CIFAR10-DVS数据集上进行实验。表4中对比了IM-loss[40],SpikeFormer[29],MPBN[38],Ternary spike[6],AGMM[33],STAA-SNN[34],FATA-SNN[39]算法在 CIFAR10-DVS 数据集的实验结果。 如表4所示,与现有国内外最优方法相比,Spiking-ReLU的准确率最高,这表明SpikingReLU可以更加有效地处理事件流数据。
SNNs使用二进制脉冲信号传递信息是其低功耗优势的主要来源。具体而言,由于二进制脉冲信号由0和1组成,具有稀疏性;其次,由于SNNs可以通过更加节能的累加操作实现。例如,在45 nm工艺下,累加操作能效为0.9 pJ,传统的乘加操作能耗为4.6 pJ。式(12)(13)为ANNs和SNN网络的功耗计算方法。
$ {E}_{\text{ANN}}={N}^{2}\cdot {C}_{\text{in}}\cdot {C}_{\text{out}}\cdot {K}^{2}\cdot {E}_{\text{MAC}}\;, $
$ {E}_{\text{SNN}}=T\cdot fr\cdot {N}^{2}\cdot {C}_{\text{in}}\cdot {C}_{\text{out}}\cdot {K}^{2}\cdot {E}_{\text{AC}}\;, $
其中N为输出特征图的尺寸,Cin为输入特征的通道数,Cout为输出特征的通道数,K为卷积核的尺寸,T为SNNs的时间步大小,fr为SNNs中脉冲信号的发射率。根据式(12)(13),分别计算了ANN(不采用脉冲驱动计算)以及SNN(采用脉冲驱动计算)的功耗。如表5所示,采用SNN可以有效地降低功耗。
本文提出了一种基于混合模型转换的高性能脉冲神经网络——SpikingReLU,该方法使用训练-推理解耦的策略,使SNNs具备高性能和事件驱动计算的优势。在训练期间,SpikingReLU将脉冲神经网络中部分神经元替换为ReLU激活层,多个结合了非线性激活层的卷积层可以显著增强脉冲神经网络的表征能力,但也破坏了其事件驱动计算的优势。针对这一问题,本文进一步提出使用重参数化技术,将包含ANNs和SNNs的高性能混合网络模型完全转换事件驱动计算。在多个主流的静态和神经形态数据集上的实验结果均充分验证了本文方法的有效性和先进性。

参考文献 引证文献
排序方式:
1
黄铁军, 余肇飞, 刘怡俊. 类脑机的思想与体系结构综述[J]. 计算机研究与发展, 2019, 56(6): 1135−1148
Huang Tiejun, Yu Zhaofei, Liu Yijun. A survey of brain-inspired machine thought and architecture[J]. Journal of Computer Research and Development, 2019, 56(6): 1135−1148 (in Chinese)
2
Guo Yufei, Huang Xuhui, Ma Zhe. Direct learning-based deep spiking neural networks: A review[J]. Frontiers in Neuroscience, 2023, 17(1): 120−795
3
Wu Yuchen, Feng Junwen. Development and application of artificial neural network[J]. Wireless Personal Communications, 2018, 102(2): 1645−1656
4
Bellec G, Salaj D, Subramoney A, et al. Long short-term memory and learning-to-learn in networks of spiking neurons[C]//Proc of the 32nd Int Conf on Neural Information Processing Systems (NeurIPS). Red Hook, NY: Curran Associates Inc, 2018: 787−797
5
Fang Wei, Yu Zhaofei, Chen Yanqi, et al. Incorporating learnable membrane time constant to enhance learning of spiking neural networks[C]//Proc of the 21st IEEE/CVF Int Conf on Computer Vision (ICCV). Los Alamitos, CA: IEEE Computer Society, 2021: 2661−2671
6
Guo Yufei, Chen Yuanpei, Liu Xiaode, et al. Ternary spike: Learning ternary spikes for spiking neural networks[C]//Proc of the 38th AAAI Conf on Artificial Intelligence. Palo Alto, CA: AAAI, 2024, 12244−12252
7
Luo Xinhao, Yao Man, Chou Yuhong, et al. Integer-valued training and spike-driven inference spiking neural network for high-performance and energy-efficient object detection[C]//Proc of the 18th European Conf on Computer Vision (ECCV). Berlin: Springer, 2024: 253−272
8
Hu Yifan, Deng Lei, Wu Yujie, et al. Advancing spiking neural networks toward deep residual learning[J]. IEEE Transactions on Neural Networks and Learning Systems, 2024, 36(2): 2353−2367
9
Su Qiaoyi, Chou Yuhong, Hu Yifan, et al. Deep directly-trained spiking neural networks for object detection[C]//Proc of the 22nd IEEE/CVF Int Conf on Computer Vision (ICCV). Los Alamitos, CA: IEEE Computer Society, 2023: 6555−6565
10
Yao Man, Zhao Guangshe, Zhang Hengyu, et al. Attention spiking neural networks[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2023, 45(8): 9393−9410
11
Yang Shu, Yu Chengting, Liu Lei, et al. Efficient ANN-guided distillation: Aligning rate-based features of spiking neural networks through hybrid block-wise replacement[C]//Proc of the 42nd IEEE/CVF Conf on Computer Vision and Pattern Recognition (CVPR). Los Alamitos, CA: IEEE Computer Society, 2025: 10025−10035
12
Yu Chengting, Zhao Xiaochen, Liu Lei, et al. Efficient logit-based knowledge distillation of deep spiking neural networks for full-range timestep deployment[C/OL]//Proc of the 42nd Int Conf on Machine Learning (ICML). 2025: 393−941. [2026-01-21]. https://openreview.net/forum?id=ZvkyeUrpsA
13
Cao Yongqiang, Chen Yang, Khosla D. Spiking deep convolutional neural networks for energy-efficient object recognition[J]. International Journal of Computer Vision, 2015, 113(1): 54−66
14
Hunsberger E, Eliasmith C. Spiking deep networks with LIF neurons[J]. arXiv preprint, arXiv: 1510.08829, 2015
15
Hunsberger E, Eliasmith C. Training spiking deep networks for neuromorphic hardware[J]. arXiv preprint, arXiv: 1611.05141, 2016
16
Diehl P U, Neil D, Binas J, et al. Fast-classifying, high-accuracy spiking deep networks through weight and threshold balancing[C]//Proc of the 2015 Int Joint Conf on Neural Networks (IJCNN). Piscataway, NJ: IEEE, 2015: 1−8
17
Zhang Anguo, Shi Jieming, Wu Junyi, et al. Low latency and sparse computing spiking neural networks with self-driven adaptive threshold plasticity[J]. IEEE Transactions on Neural Networks and Learning Systems, 2024, 35(12): 17177−17188
18
Bu Tong, Ding Jianhao, Yu Zhaofei, et al. Optimized potential initialization for low-latency spiking neural networks[C]//Proc of the 36th AAAI Conf on Artificial Intelligence. Palo Alto, CA: AAAI, 2022, 11−20
19
Yao Xingting, Li Fanrong, Mo Zitao, et al. GLIF: A unified gated leaky integrate-and-fire neuron for spiking neural networks[C]//Proc of the 36th Int Conf on Neural Information Processing Systems (NeurIPS). San Diego, CA: Neural Information Processing Systems Foundation, Inc, 2022: 32160−32171
20
Guo Yufei, Liu Xiaode, Chen Yuanpei, et al. Spiking Transformer: introducing accurate addition-only spiking self-attention for transformer[C]//Proc of the 42nd IEEE/CVF Conf on Computer Vision and Pattern Recognition (CVPR). Los Alamitos, CA: IEEE Computer Society, 2025: 24398−24408
21
Guo Yufei, Zhang Liwen, Chen Yuanpei, et al. Real spike: Learning real-valued spikes for spiking neural networks[C]//Proc of the 17th European Conf on Computer Vision (ECCV). Berlin: Springer, 2022: 52−68
22
冯烁, 路冬冬, 尹飞, 等. 面向分布式计算的类脑智能处理器指令集设计[J]. 计算机研究与发展, 2025, 62(1): 1−14
Feng Shuo, Lu Dongdong, Yin Fei, et al. Instruction set design for brain-inspired intelligent processors oriented to distributed computing[J]. Journal of Computer Research and Development, 2025, 62(1): 1−14 (in Chinese)
23
杨智杰, 王蕾, 石伟, 等. 类脑处理器异步片上网络架构[J]. 计算机研究与发展, 2023, 60(1): 17−29
Yang Zhijie, Wang Lei, Shi Wei, et al. Asynchronous network-on-chip architecture for brain-inspired processors[J]. Journal of Computer Research and Development, 2023, 60(1): 17−29 (in Chinese)
24
王智超, 陈亮, 李千鹏, 等. 基于混合模式路由的脑启发片上网络架构[J]. 计算机研究与发展, 2025, 62(11): 2663−2676
Wang Zhichao, Chen Liang, Li Qianpeng, et al. A brain-inspired network-on-chip architecture with hybrid-mode routing[J]. Journal of Computer Research and Development, 2025, 62(11): 2663−2676 (in Chinese)
25
Sengupta A, Ye Yuting, Wang R, et al. Going deeper in spiking neural networks: VGG and residual architectures[J]. Frontiers in Neuroscience, 2019, 13(1): 95−106
26
Han Bing, Srinivasan G, Roy K. RMP-SNN: Residual membrane potential neuron for enabling deeper high-accuracy and low-latency spiking neural network[C]//Proc of the 37th IEEE/CVF Conf on Computer Vision and Pattern Recognition (CVPR). Los Alamitos, CA: IEEE Computer Society, 2020: 13558−13567
27
Yoo D, Jeong D S. Cbp-Qsnn: Spiking neural networks quantized using constrained backpropagation[J]. IEEE Journal on Emerging and Selected Topics in Circuits and Systems, 2023, 13(4): 1137−1146
28
Meng Qingyan, Xiao Mingqing, Yan Shen, et al. Training high-performance low-latency spiking neural networks by differentiation on spike representation[C]//Proc of the 39th IEEE/CVF Conf on Computer Vision and Pattern Recognition (CVPR). Los Alamitos, CA: IEEE Computer Society, 2022: 12444−12453
29
Zhou Zhaokun, Zhu Yuesheng, He Chao, et al. SpikFormer: When spiking neural network meets transformer[C/OL]//Proc of the 11th Int Conf on Learning Representations (ICLR). 2023[2026-01-21]. https://openreview.net/forum?id=frE4fUwz_h
30
Yao Man, Hu Jiakui, Zhou Zhaokun, et al. Spike-driven Transformer[C]//Proc of the 37th Int Conf on Neural Information Processing Systems (NeurIPS). New Orleans, CA: Curran Associates, 2023: 64043−64058
31
Fang Yi, Wang Zhaokun, Zhang Lei, et al. Spiking wavelet Transformer[C]//Proc of the 18th European Conf on Computer Vision (ECCV). Berlin: Springer, 2024: 19−37
32
Wang Shu, Zhang Ming, Zhang Di, et al. Spiking vision Transformer with saccadic attention[C/OL]//Proc of the Int Conf on Learning Representations (ICLR). 2025[2026-01-21]. https://openreview.net/forum?id=qzZsz6MuEq
33
Liang Yixing, Wei Wenjie, Belatreche A, et al. Towards accurate binary spiking neural networks: Learning with adaptive gradient modulation mechanism[C]//Proc of the 39th AAAI Conf on Artificial Intelligence. Palo Alto, CA: AAAI, 2025, 39(2): 1402−1410
34
Zhang Tianpeng, Yu Kexin, Zhong Xinghao, et al. STAA-SNN: Spatial-temporal attention aggregator for spiking neural networks[C]//Proc of the 42nd IEEE/CVF Conf on Computer Vision and Pattern Recognition (CVPR). Los Alamitos, CA: IEEE Computer Society, 2025: 13959−13969
35
Deng Shikuang, Li Yang, Zhang Shiyan, et al. Temporal efficient training of spiking neural network via gradient re-weighting[C/OL]//Proc of the Int Conf on Learning Representations (ICLR). 2022[2026-01-21]. https://openreview.net/forum?id=_XNtisL32jv
36
Guo Yufei, Tong Xinyi, Chen Yuanpei, et al. RecDis-SNN: Rectifying membrane potential distribution for directly training spiking neural networks[C]//Proc of the 39th IEEE/CVF Conf on Computer Vision and Pattern Recognition (CVPR). Los Alamitos, CA: IEEE Computer Society, 2022: 326−335
37
Guo Yufei, Liu Xiaode, Chen Yuanpei, et al. RMP-loss: Regularizing membrane potential distribution for spiking neural networks[C]//Proc of the 22nd IEEE/CVF Int Conf on Computer Vision (ICCV). Los Alamitos, CA: IEEE Computer Society, 2023: 17391−17401
38
Guo Yufei, Zhang Yuhan, Chen Yuanpei, et al. Membrane potential batch normalization for spiking neural networks[C]//Proc of the 22nd IEEE/CVF Int Conf on Computer Vision (ICCV). Los Alamitos, CA: IEEE Computer Society, 2023: 19420−19430
39
Yu Kexin, Zhang Tianpeng, Wang Hao, et al. FSTA-SNN: Frequency-based spatial-temporal attention module for spiking neural networks[C]//Proc of the 39th AAAI Conf on Artificial Intelligence. Palo Alto, CA: AAAI, 2025: 22227−22235
40
Guo Yufei, Chen Yuanpei, Zhang Liwen, et al. IM-loss: Information maximization loss for spiking neural networks[C]//Proc of the 36th Int Conf on Neural Information Processing Systems (NeurIPS). San Diego, CA: Neural Information Processing Systems Foundation, Inc, 2022: 156−166
2026年第63卷第8期
75
33
0
引用本文
BibTeX
文章信息
doi: 10.7544/issn1000-1239.202550722
  • 接收时间:2025-09-30
  • 首发时间:2026-11-17
  • 出版时间:2026-08-01
补充材料
相关文章
文章信息
作者
出版历史
  • 收稿日期:2025-09-30
  • 修回日期:2026-03-30
基金
作者信息
    1启元实验室 北京 100095
    2军事科学院国防科技创新研究院 北京 100091
    3北京跟踪与通信技术研究所 北京 100094
    4陕西省信息获取与处理重点实验室(西北工业大学) 西安 710129

通讯作者:

参考文献
分享链接
https://castjournals.cast.org.cn/joweb/jsjyjyfz/CN/10.7544/issn1000-1239.202550722
分享至
全文二维码

扫描看全文

引用本文
BibTeX
本文的引用情况
2种不同金属材料的力学参数
科
Family
属数
Number of
genus
种数
Number of
species
占总种数比例
Percentage of
total species (%)
属
Genus
种数
Number of
species
占总种数比例
Percentage of total
species (%)
鹅膏菌科Amanitaceae 2 11 5.26 鹅膏菌属 Amanita 10 4.78
小菇科 Mycenaceae 2 12 5.74 丝盖伞属 Inocybe 5 2.39
多孔菌科 Polyporaceae 8 14 6.70 蜡蘑属 Laccaria 5 2.39
红菇科 Russulaceae 3 23 11.00 小皮伞属 Marasmius 6 2.87
小菇属 Mycena 11 5.26
光柄菇属 Pluteus 5 2.39
红菇属 Russula 17 8.13
栓菌属 Trametes 5 2.39
关闭全屏