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.

BSNN: Towards faster and better conversion of artificial neural networks to spiking neural networks with bistable neurons

BitSNNs: Revisiting Energy-efficient Spiking Neural Networks

Spiking Deep Residual Networks.

Faster and Stronger: When ANN-SNN Conversion Meets Parallel Spiking Calculation

Efficient and Accurate Conversion of Spiking Neural Network with Burst Spikes

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

SpikeConverter: an Efficient Conversion Framework Zipping the Gap Between Artificial Neural Networks and Spiking Neural Networks

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

Optimal Conversion of Conventional Artificial Neural Networks to Spiking Neural Networks

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

Optimized Potential Initialization for Low-latency Spiking Neural Networks

Optimal ANN-SNN Conversion for Fast and Accurate Inference in Deep Spiking Neural Networks

Symmetric-threshold ReLU for Fast and Nearly Lossless ANN-SNN Conversion

Optimal ANN-SNN Conversion for High-accuracy and Ultra-low-latency Spiking Neural Networks

MSAT: Biologically Inspired Multi-Stage Adaptive Threshold for Conversion of Spiking Neural Networks

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

Converting High-Performance and Low-Latency SNNs through Explicit Modelling of Residual Error in ANNs

Spike Calibration: Bridging the Gap between ANNs and SNNs in ANN-SNN Conversion

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

Optimizing Power Efficiency: Converting Recurrent Neural Networks to Spiking Neural Networks for Time-Domain Analysis

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