Skip to main content

ORIGINAL RESEARCH article

Front. Neurosci.
Sec. Neuromorphic Engineering
Volume 18 - 2024 | doi: 10.3389/fnins.2024.1449020
This article is part of the Research Topic Efficient AI processing (computing) and adaptation in Neuromorphic systems View all articles

An All Integer-based Spiking Neural Network with Dynamic Threshold Adaptation

Provisionally accepted
Chenglong Zou Chenglong Zou 1,2*Xiaoxin Cui Xiaoxin Cui 3Yi Zhong Yi Zhong 3Zhenhui Dai Zhenhui Dai 3Yuan WANG Yuan WANG 3
  • 1 School of Mathematical Science, Peking University, Beijing, China
  • 2 Peking University Chongqing Research Institute of Big Data, Chongqing, China
  • 3 School of Integrated Circuits, Peking University, Beijing, China

The final, formatted version of the article will be published soon.

    Spiking Neural Networks (SNNs) are typically regards as the third generation of neural networks due to their inherent event-driven computing capabilities and remarkable energy efficiency. However, training an SNN that possesses fast inference speed and comparable accuracy to modern artificial neural networks (ANNs) remains a considerable challenge. In this paper, a sophisticated SNN modeling algorithm incorporating a novel dynamic threshold adaptation mechanism is proposed. It aims to eliminate the spiking synchronization error commonly occurred in many traditional ANN2SNN conversion works. Additionally, all variables in the proposed SNNs, including the membrane potential, threshold and synaptic weights, are quantized to integers, making them highly compatible with hardware implementation. Experimental results indicate that the proposed spiking LeNet and VGG-Net achieve accuracies exceeding 99.45% and 93.15% on the MNIST and CIFAR-10 datasets, respectively, with only 4 and 8 time steps required for simulating one sample. Due to this all integerbased quantization process, the required computational operations are significantly reduced, potentially providing a substantial energy efficiency advantage for numerous edge computing applications.

    Keywords: Spiking Neural network, Dynamic threshold adaptation, ANN2SNN conversion, Network quantization, neuromorphic computing

    Received: 14 Jun 2024; Accepted: 19 Nov 2024.

    Copyright: © 2024 Zou, Cui, Zhong, Dai and WANG. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

    * Correspondence: Chenglong Zou, School of Mathematical Science, Peking University, Beijing, China

    Disclaimer: All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.