CompSNN: A lightweight spiking neural network based on spatiotemporally compressive spike features

doi:10.1016/j.neucom.2020.10.100

	CompSNN: A lightweight spiking neural network based on spatiotemporally compressive spike features
	Wang, Tengxiao 2; Shi, Cong 2; Zhou, Xichuan 2; Lin, Yingcheng 2; He, Junxian 2; Gan, Ping 2; Li, Ping 2; Wang, Ying 3; Liu, Liyuan 1; Wu, Nanjian 1
刊名	NEUROCOMPUTING
	2021-02-15
卷号	425 页码:96-106
关键词	Neuromorphic computing Spiking neural networks SNN Compressive sensing Object classification Multi-spike encoding
ISSN号	0925-2312
DOI	10.1016/j.neucom.2020.10.100
英文摘要	Brain-inspired spiking neural networks (SNNs) have become a research hotspot in recent years. These SNNs communicate and process information in a form of spatiotemporally sparse spikes, leading to high energy efficiency and low computational cost for object classification tasks. However, to reduce computational complexity while maintaining SNN classification accuracy still remains a challenge. Extracting representative and robust feature is the key. This paper proposes efficient spatiotemporally compressive spike features and presents a lightweight SNN framework that includes a feature extraction layer to extract such compressive features. Our experiments based on popular benchmark datasets demonstrated that the spatiotemporally compressive spike features are competent and robust in representing the input spike trains. The experimental results also suggest that our lightweight SNN framework with such compressive spike feature requires a small amount of processing time consumption while achieving comparable classification rate across many popular datasets: MNIST, MNIST-DVS, Poker-DVS, Posture-DVS and more challenging Fashion-MNIST datasets. The SNN framework has a potential to be applied in low-cost or resource-limited edge computing systems and embedded devices. ? 2020 Elsevier B.V. All rights reserved.
资助项目	Key Project of Chongqing Science and Technology Foundation[cstc2019jcyjzdxmX0017] ; State Key Laboratory of Computer Architecture, Institute of Computing Technology, Chinese Academy of Sciences[CARCH201908] ; Fundamental Research Funds for the Central Universities[2019CDXYTX0024]
WOS研究方向	Computer Science
语种	英语
出版者	ELSEVIER
WOS记录号	WOS:000632015900008
内容类型	期刊论文
源URL	[http://119.78.100.204/handle/2XEOYT63/16763]
专题	中国科学院计算技术研究所
通讯作者	Shi, Cong
作者单位	1.Chinese Acad Sci, Inst Semicond, State Key Lab Superlattices & Microstruct, Beijing 100083, Peoples R China 2.Chongqing Univ, Sch Microelect & Commun Engn, Chongqing 400044, Peoples R China 3.Chinese Acad Sci, Inst Comp Technol, State Key Lab Comp Architecture, Beijing 100190, Peoples R China 4.Harvard Med Sch, Schepens Eye Res Inst, Dept Ophthalmol, Massachusetts Eye & Ear, Boston, MA 02114 USA
推荐引用方式 GB/T 7714	Wang, Tengxiao,Shi, Cong,Zhou, Xichuan,et al. CompSNN: A lightweight spiking neural network based on spatiotemporally compressive spike features[J]. NEUROCOMPUTING,2021,425:96-106.
APA	Wang, Tengxiao.,Shi, Cong.,Zhou, Xichuan.,Lin, Yingcheng.,He, Junxian.,...&Luo, Gang.(2021).CompSNN: A lightweight spiking neural network based on spatiotemporally compressive spike features.NEUROCOMPUTING,425,96-106.
MLA	Wang, Tengxiao,et al."CompSNN: A lightweight spiking neural network based on spatiotemporally compressive spike features".NEUROCOMPUTING 425(2021):96-106.