Optimized spiking neurons can classify images with high accuracy through temporal coding with two spikes

尖峰神经网络 计算机科学 神经形态工程学 人工智能 Spike(软件开发) 人工神经网络 深度学习 深层神经网络 模式识别(心理学) 编码(社会科学) 数学 统计 软件工程
作者
Christoph Stöckl,Wolfgang Maass
出处
期刊:Nature Machine Intelligence [Nature Portfolio]
卷期号:3 (3): 230-238 被引量:69
标识
DOI:10.1038/s42256-021-00311-4
摘要

Spike-based neuromorphic hardware promises to reduce the energy consumption of image classification and other deep-learning applications, particularly on mobile phones and other edge devices. However, direct training of deep spiking neural networks is difficult, and previous methods for converting trained artificial neural networks to spiking neurons were inefficient because the neurons had to emit too many spikes. We show that a substantially more efficient conversion arises when one optimizes the spiking neuron model for that purpose, so that it not only matters for information transmission how many spikes a neuron emits, but also when it emits those spikes. This advances the accuracy that can be achieved for image classification with spiking neurons, and the resulting networks need on average just two spikes per neuron for classifying an image. In addition, our new conversion method improves latency and throughput of the resulting spiking networks. Spiking neural networks could offer a low-energy consuming solution to deep learning applications on the edge and in mobile devices. Using temporal coding, where the timing of spikes carries extra information, a new method efficiently converts conventional artificial neural networks to spiking networks.
最长约 10秒,即可获得该文献文件

科研通智能强力驱动
Strongly Powered by AbleSci AI
更新
PDF的下载单位、IP信息已删除 (2025-6-4)

科研通是完全免费的文献互助平台,具备全网最快的应助速度,最高的求助完成率。 对每一个文献求助,科研通都将尽心尽力,给求助人一个满意的交代。
实时播报
老Mark完成签到,获得积分10
刚刚
刚刚
711发布了新的文献求助10
刚刚
等等完成签到,获得积分10
刚刚
饱满友易发布了新的文献求助10
1秒前
hellosci666完成签到,获得积分10
1秒前
2秒前
所所应助wuniuniu采纳,获得10
3秒前
但愿所及完成签到,获得积分10
3秒前
3秒前
5秒前
小北完成签到,获得积分10
5秒前
ziwei完成签到,获得积分10
5秒前
怡然云朵发布了新的文献求助10
5秒前
瘦瘦谷兰完成签到 ,获得积分10
5秒前
尔安完成签到,获得积分10
6秒前
柒月小鱼完成签到 ,获得积分10
6秒前
shuangma发布了新的文献求助10
6秒前
Lucas应助泽1采纳,获得10
6秒前
Jayce完成签到,获得积分10
7秒前
Mannstop发布了新的文献求助10
7秒前
7秒前
失眠的汽车完成签到,获得积分10
7秒前
7秒前
山色青完成签到,获得积分10
7秒前
核动力驴完成签到,获得积分10
8秒前
机智的三国菌完成签到,获得积分10
8秒前
9秒前
逗我的人继续逗我完成签到,获得积分10
9秒前
木木木木完成签到,获得积分10
9秒前
9秒前
wwzj完成签到,获得积分10
10秒前
Jayce发布了新的文献求助10
10秒前
传奇3应助ahhhh采纳,获得10
11秒前
11秒前
胡图图完成签到 ,获得积分10
12秒前
嚭嚭完成签到,获得积分10
12秒前
研友_8yN60L完成签到,获得积分10
12秒前
12秒前
13秒前
高分求助中
The Mother of All Tableaux Order, Equivalence, and Geometry in the Large-scale Structure of Optimality Theory 2400
Ophthalmic Equipment Market by Devices(surgical: vitreorentinal,IOLs,OVDs,contact lens,RGP lens,backflush,diagnostic&monitoring:OCT,actorefractor,keratometer,tonometer,ophthalmoscpe,OVD), End User,Buying Criteria-Global Forecast to2029 2000
A new approach to the extrapolation of accelerated life test data 1000
Cognitive Neuroscience: The Biology of the Mind (Sixth Edition) 1000
Official Methods of Analysis of AOAC INTERNATIONAL 600
ACSM’s Guidelines for Exercise Testing and Prescription, 12th edition 588
A Preliminary Study on Correlation Between Independent Components of Facial Thermal Images and Subjective Assessment of Chronic Stress 500
热门求助领域 (近24小时)
化学 材料科学 医学 生物 工程类 有机化学 生物化学 物理 内科学 纳米技术 计算机科学 化学工程 复合材料 遗传学 基因 物理化学 催化作用 冶金 细胞生物学 免疫学
热门帖子
关注 科研通微信公众号,转发送积分 3960387
求助须知:如何正确求助?哪些是违规求助? 3506503
关于积分的说明 11130906
捐赠科研通 3238717
什么是DOI,文献DOI怎么找? 1789884
邀请新用户注册赠送积分活动 871982
科研通“疑难数据库(出版商)”最低求助积分说明 803118