Spiking Deep Residual Network

TL;DR

This paper introduces a novel method to convert deep residual neural networks into spiking neural networks, achieving high performance on large-scale datasets and enabling deeper SNNs than previously possible.

Contribution

It presents a new approach for building deep residual spiking neural networks by converting trained ResNets, including a shortcut conversion model and error compensation mechanism.

Findings

Achieved state-of-the-art performance on CIFAR-10, CIFAR-100, and ImageNet.

First SNN deeper than 40 layers with performance comparable to ANNs.

Demonstrated effective conversion method maintaining accuracy in deep SNNs.

Abstract

Spiking neural networks (SNNs) have received significant attention for their biological plausibility. SNNs theoretically have at least the same computational power as traditional artificial neural networks (ANNs). They possess potential of achieving energy-efficiency while keeping comparable performance to deep neural networks (DNNs). However, it is still a big challenge to train a very deep SNN. In this paper, we propose an efficient approach to build a spiking version of deep residual network (ResNet). ResNet is considered as a kind of the state-of-the-art convolutional neural networks (CNNs). We employ the idea of converting a trained ResNet to a network of spiking neurons, named Spiking ResNet (S-ResNet). We propose a shortcut conversion model to appropriately scale continuous-valued activations to match firing rates in SNN, and a compensation mechanism to reduce the error caused by discretisation. Experimental results demonstrate that, compared with the state-of-the-art SNN approaches, the proposed Spiking ResNet achieves the best performance on CIFAR-10, CIFAR-100, and ImageNet 2012. Our work is the first time to build a SNN deeper than 40, with comparable performance to ANNs on a large-scale dataset.