Abstract:Spiking Neural Networks (SNNs) are a pathway that could potentially empower low-power event-driven neuromorphic hardware due to their spatio-temporal information processing capability and high biological plausibility. Although SNNs are currently more efficient than artificial neural networks (ANNs), they are not as accurate as ANNs. Error backpropagation is the most common method for directly training neural networks, promoting the prosperity of ANNs in various deep learning fields. However, since the signals transmitted in the SNN are non-differentiable discrete binary spike events, the activation function in the form of spikes presents difficulties for the gradient-based optimization algorithms to be directly applied in SNNs, leading to a performance gap (i.e., accuracy and latency) between SNNs and ANNs. This paper introduces a new learning algorithm, called SSTDP, which bridges the gap between backpropagation (BP)-based learning and spike-time-dependent plasticity (STDP)-based learning to train SNNs efficiently. The scheme incorporates the global optimization process from BP and the efficient weight update derived from STDP. It not only avoids the non-differentiable derivation in the BP process but also utilizes the local feature extraction property of STDP. Consequently, our method can lower the possibility of vanishing spikes in BP training and reduce the number of time steps to reduce network latency. In SSTDP, we employ temporal-based coding and use Integrate-and-Fire (IF) neuron as the neuron model to provide considerable computational benefits. Our experiments show the effectiveness of the proposed SSTDP learning algorithm on the SNN by achieving the best classification accuracy 99.3% on the Caltech 101 dataset, 98.1% on the MNIST dataset, and 91.3% on the CIFAR-10 dataset compared to other SNNs trained with other learning methods. It also surpasses the best inference accuracy of the directly trained SNN with 25~32× less inference latency. Moreover, we analyze event-based computations to demonstrate the efficacy of the SNN for inference operation in the spiking domain, and SSTDP methods can achieve 1.3~37.7× fewer addition operations per inference. The code is available at: https://github.com/MXHX7199/SNN-SSTDP .

Online Pseudo-Zeroth-Order Training of Neuromorphic Spiking Neural Networks

Online Training Through Time for Spiking Neural Networks

A High Energy-Efficiency Multi-core Neuromorphic Architecture for Deep SNN Training

Mapping Very Large Scale Spiking Neuron Network to Neuromorphic Hardware.

Energy Efficient Training of SNN Using Local Zeroth Order Method

A Neuromorphic-Hardware Oriented Bio-Plausible Online-Learning Spiking Neural Network Model.

Online Stabilization of Spiking Neural Networks.

Comprehensive Online Training and Deployment for Spiking Neural Networks

A A 22nm 0.43pJ/SOP Sparsity-Aware In-Memory Neuromorphic Computing System with Hybrid Spiking and Artificial Neural Network and Configurable Topology

In-Hardware Learning of Multilayer Spiking Neural Networks on a Neuromorphic Processor

Towards Memory- and Time-Efficient Backpropagation for Training Spiking Neural Networks

Neuromorphic Online Learning for Spatiotemporal Patterns with a Forward-only Timeline

The Backpropagation Algorithm Implemented on Spiking Neuromorphic Hardware

A temporally and spatially local spike-based backpropagation algorithm to enable training in hardware

An Efficient Non-Backpropagation Method for Training Spiking Neural Networks

Scalable NoC-based Neuromorphic Hardware Learning and Inference

BrainScale: Enabling Scalable Online Learning in Spiking Neural Networks

Efficient Implementation of Spiking Neural Networks for Inference Using Ex-Situ Training

In-situ Learning in Hardware Compatible Multi-layer Memristive Spiking Neural Network

SSTDP: Supervised Spike Timing Dependent Plasticity for Efficient Spiking Neural Network Training

TripleBrain: An Edge Neuromorphic Architecture for High-accuracy Single-layer Spiking Neural Network with On-chip Self-organizing and Reinforcement Learning