Abstract:Spiking neural networks (SNNs) can do inference with low power consumption due to their spike sparsity. Although SNNs can be combined with neuromorphic hardware to achieve efficient inference, they are often difficult to train directly due to discrete non-differentiable spikes. As an alternative, ANN-SNN conversion is an efficient way to achieve deep SNNs by converting well-trained artificial neural networks (ANNs). However, the existing methods commonly use constant threshold for conversion. A high constant threshold value prevents neurons from rapidly delivering spikes to deeper layers and causes high time delay. In addition, the same response for different inputs may result in information loss during the information transmission. Inspired by the biological adaptive threshold mechanism, we propose a multistage adaptive threshold (MSAT) method to alleviate this problem. Instead of using a single, constant value, the threshold is adjusted in multistages, adapting to each neuron’s firing history and input properties. Specifically, for each neuron, the dynamic threshold is positively correlated with the average membrane potential and negatively correlated with the rate of depolarization. The adaptation to membrane potential and input allows a timely adjustment of the threshold to fire spikes faster and transmit more information. Moreover, we analyze the spikes of inactivated neurons error, which is pervasive in early time steps. We also propose spike confidence accordingly to measure confidence about the neurons that correctly deliver spikes. Such spike confidence in early time steps is used to determine whether to elicit the spike to alleviate the spikes of inactivated neurons error. Combined with the proposed methods, we examine the performance on CIFAR-10, CIFAR-100, and ImageNet datasets. We also conduct sentiment classification and speech recognition experiments on the IDBM and Google speech commands datasets, respectively. Experiments show that our methods can achieve near-lossless and lower latency ANN-SNN conversion. In summary, we build a biologically inspired multistage adaptive threshold for converted SNN, with comparable performance to state-of-the-art methods while improving energy efficiency.

Deep Spiking Neural Networks with Binary Weights for Object Recognition

Efficient Structure Slimming for Spiking Neural Networks

Spike Trains Encoding and Threshold Rescaling Method for Deep Spiking Neural Networks

Deep Spiking Binary Neural Network for Digital Neuromorphic Hardware

Spiking Deep Convolutional Neural Networks for Energy-Efficient Object Recognition

Deep Spiking Convolutional Neural Networks for Programmable Neuro-synaptic System

Spiking Deep Residual Networks.

ALBSNN: ultra-low latency adaptive local binary spiking neural network with accuracy loss estimator

Optimal Conversion of Conventional Artificial Neural Networks to Spiking Neural Networks

Toward a Lossless Conversion for Spiking Neural Networks with Negative-Spike Dynamics.

Toward High-Accuracy and Low-Latency Spiking Neural Networks With Two-Stage Optimization

Enabling Deep Spiking Neural Networks with Hybrid Conversion and Spike Timing Dependent Backpropagation

Low Latency Spiking ConvNets with Restricted Output Training and False Spike Inhibition

Direct Training of Hardware-Friendly Weight Binarized Spiking Neural Network with Surrogate Gradient Learning Towards Spatio-Temporal Event-Based Dynamic Data Recognition

Efficient and Accurate Conversion of Spiking Neural Network with Burst Spikes

Towards Lossless ANN-SNN Conversion under Ultra-Low Latency with Dual-Phase Optimization

Ultra-low Latency Adaptive Local Binary Spiking Neural Network with Accuracy Loss Estimator

MSAT: biologically inspired multistage adaptive threshold for conversion of spiking neural networks

Spike Calibration: Fast and Accurate Conversion of Spiking Neural Network for Object Detection and Segmentation

A universal ANN-to-SNN framework for achieving high accuracy and low latency deep Spiking Neural Networks

Deep Spiking Neural Network with Ternary Spikes