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

Yijian Pei; Changqing Xu; Zili Wu; Yi Liu; Yintang Yang

doi:10.3389/fnins.2023.1225871

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

Front Neurosci. 2023 Sep 13:17:1225871. doi: 10.3389/fnins.2023.1225871. eCollection 2023.

Authors

Yijian Pei^#¹, Changqing Xu^#^{1

2}, Zili Wu³, Yi Liu², Yintang Yang²

Affiliations

¹ Guangzhou Institute of Technology, Xidian University, Xi'an, China.
² School of Microelectronics, Xidian University, Xi'an, China.
³ School of Computer Science and Technology, Xidian University, Xi'an, China.

^# Contributed equally.

Abstract

Spiking neural network (SNN) is a brain-inspired model with more spatio-temporal information processing capacity and computational energy efficiency. However, with the increasing depth of SNNs, the memory problem caused by the weights of SNNs has gradually attracted attention. In this study, we propose an ultra-low latency adaptive local binary spiking neural network (ALBSNN) with accuracy loss estimators, which dynamically selects the network layers to be binarized to ensure a balance between quantization degree and classification accuracy by evaluating the error caused by the binarized weights during the network learning process. At the same time, to accelerate the training speed of the network, the global average pooling (GAP) layer is introduced to replace the fully connected layers by combining convolution and pooling. Finally, to further reduce the error caused by the binary weight, we propose binary weight optimization (BWO), which updates the overall weight by directly adjusting the binary weight. This method further reduces the loss of the network that reaches the training bottleneck. The combination of the above methods balances the network's quantization and recognition ability, enabling the network to maintain the recognition capability equivalent to the full precision network and reduce the storage space by more than 20%. So, SNNs can use a small number of time steps to obtain better recognition accuracy. In the extreme case of using only a one-time step, we still can achieve 93.39, 92.12, and 69.55% testing accuracy on three traditional static datasets, Fashion- MNIST, CIFAR-10, and CIFAR-100, respectively. At the same time, we evaluate our method on neuromorphic N-MNIST, CIFAR10-DVS, and IBM DVS128 Gesture datasets and achieve advanced accuracy in SNN with binary weights. Our network has greater advantages in terms of storage resources and training time.

Keywords: binary neural networks; neuromorphic computing; sparsity; spiking neural networks; visual recognition.

Grants and funding

This study was supported by the National Natural Science Foundation of China under Grant 62004146, by the China Postdoctoral Science Foundation funded project under Grant 2021M692498, by the Fundamental Research Funds for the Central Universities under Grant XJSJ23106, and by Science and Technology Projects in Guangzhou under Grant SL2022A04J00095.