Efficiently Training Time-to-First-Spike Spiking Neural Networks from Scratch

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Che, Kaiwei, Fang, Wei, Ma, Zhengyu, Huang, Yifan, Xue, Peng, Yuan, Li, Masquelier, Timothée, Tian, Yonghong
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866913726542643200
author Che, Kaiwei
Fang, Wei
Ma, Zhengyu
Huang, Yifan
Xue, Peng
Yuan, Li
Masquelier, Timothée
Tian, Yonghong
author_facet Che, Kaiwei
Fang, Wei
Ma, Zhengyu
Huang, Yifan
Xue, Peng
Yuan, Li
Masquelier, Timothée
Tian, Yonghong
contents Spiking Neural Networks (SNNs), with their event-driven and biologically inspired operation, are well-suited for energy-efficient neuromorphic hardware. Neural coding, critical to SNNs, determines how information is represented via spikes. Time-to-First-Spike (TTFS) coding, which uses a single spike per neuron, offers extreme sparsity and energy efficiency but suffers from unstable training and low accuracy due to its sparse firing. To address these challenges, we propose a training framework incorporating parameter initialization, training normalization, temporal output decoding, and pooling layer re-evaluation. The proposed parameter initialization and training normalization mitigate signal diminishing and gradient vanishing to stabilize training. The output decoding method aggregates temporal spikes to encourage earlier firing, thereby reducing the latency. The re-evaluation of the pooling layer indicates that average-pooling keeps the single-spike characteristic and that max-pooling should be avoided. Experiments show the framework stabilizes and accelerates training, reduces latency, and achieves state-of-the-art accuracy for TTFS SNNs on MNIST (99.48%), Fashion-MNIST (92.90%), CIFAR10 (90.56%), and DVS Gesture (95.83%).
format Preprint
id arxiv_https___arxiv_org_abs_2410_23619
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Efficiently Training Time-to-First-Spike Spiking Neural Networks from Scratch
Che, Kaiwei
Fang, Wei
Ma, Zhengyu
Huang, Yifan
Xue, Peng
Yuan, Li
Masquelier, Timothée
Tian, Yonghong
Neural and Evolutionary Computing
Spiking Neural Networks (SNNs), with their event-driven and biologically inspired operation, are well-suited for energy-efficient neuromorphic hardware. Neural coding, critical to SNNs, determines how information is represented via spikes. Time-to-First-Spike (TTFS) coding, which uses a single spike per neuron, offers extreme sparsity and energy efficiency but suffers from unstable training and low accuracy due to its sparse firing. To address these challenges, we propose a training framework incorporating parameter initialization, training normalization, temporal output decoding, and pooling layer re-evaluation. The proposed parameter initialization and training normalization mitigate signal diminishing and gradient vanishing to stabilize training. The output decoding method aggregates temporal spikes to encourage earlier firing, thereby reducing the latency. The re-evaluation of the pooling layer indicates that average-pooling keeps the single-spike characteristic and that max-pooling should be avoided. Experiments show the framework stabilizes and accelerates training, reduces latency, and achieves state-of-the-art accuracy for TTFS SNNs on MNIST (99.48%), Fashion-MNIST (92.90%), CIFAR10 (90.56%), and DVS Gesture (95.83%).
title Efficiently Training Time-to-First-Spike Spiking Neural Networks from Scratch
topic Neural and Evolutionary Computing
url https://arxiv.org/abs/2410.23619