Spike Accumulation Forwarding for Effective Training of Spiking Neural Networks

TL;DR

This paper introduces Spike Accumulation Forwarding (SAF), a novel training paradigm for spiking neural networks that reduces computational complexity and memory usage while maintaining accuracy, with theoretical validation and experimental confirmation.

Contribution

SAF is a new training method for SNNs that halves forward operations, is theoretically consistent with Spike Representation and OTTT, and improves efficiency without sacrificing accuracy.

Findings

SAF halves the number of operations during forward passes.

SAF reduces memory and training time.

SAF maintains accuracy comparable to existing methods.

Abstract

In this article, we propose a new paradigm for training spiking neural networks (SNNs), spike accumulation forwarding (SAF). It is known that SNNs are energy-efficient but difficult to train. Consequently, many researchers have proposed various methods to solve this problem, among which online training through time (OTTT) is a method that allows inferring at each time step while suppressing the memory cost. However, to compute efficiently on GPUs, OTTT requires operations with spike trains and weighted summation of spike trains during forwarding. In addition, OTTT has shown a relationship with the Spike Representation, an alternative training method, though theoretical agreement with Spike Representation has yet to be proven. Our proposed method can solve these problems; namely, SAF can halve the number of operations during the forward process, and it can be theoretically proven that SAF is consistent with the Spike Representation and OTTT, respectively. Furthermore, we confirmed the above contents through experiments and showed that it is possible to reduce memory and training time while maintaining accuracy.