Saved in:
Bibliographic Details
Main Authors: Xu, Changqing, Sun, Guoqing, Liu, Yi, Liao, Xinfang, Yang, Yintang
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2508.01223
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866913972018479104
author Xu, Changqing
Sun, Guoqing
Liu, Yi
Liao, Xinfang
Yang, Yintang
author_facet Xu, Changqing
Sun, Guoqing
Liu, Yi
Liao, Xinfang
Yang, Yintang
contents Reversible Spiking Neural Networks (RevSNNs) enable memory-efficient training by reconstructing forward activations during backpropagation, but suffer from high latency due to strictly sequential computation. To overcome this limitation, we propose ParaRevSNN, a parallel reversible SNN architecture that decouples sequential dependencies between reversible blocks while preserving reversibility. This design enables inter-block parallelism, significantly accelerating training and inference while retaining the memory-saving benefits of reversibility. Experiments on CIFAR10, CIFAR100, CIFAR10-DVS, and DVS128 Gesture demonstrate that ParaRevSNN matches or exceeds the accuracy of standard RevSNNs, while reducing training time by up to 35.2\% and inference time to 18.15\%, making it well-suited for deployment in resource-constrained scenarios.
format Preprint
id arxiv_https___arxiv_org_abs_2508_01223
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle ParaRevSNN: A Parallel Reversible Spiking Neural Network for Efficient Training and Inference
Xu, Changqing
Sun, Guoqing
Liu, Yi
Liao, Xinfang
Yang, Yintang
Computer Vision and Pattern Recognition
68T10
I.4.6
Reversible Spiking Neural Networks (RevSNNs) enable memory-efficient training by reconstructing forward activations during backpropagation, but suffer from high latency due to strictly sequential computation. To overcome this limitation, we propose ParaRevSNN, a parallel reversible SNN architecture that decouples sequential dependencies between reversible blocks while preserving reversibility. This design enables inter-block parallelism, significantly accelerating training and inference while retaining the memory-saving benefits of reversibility. Experiments on CIFAR10, CIFAR100, CIFAR10-DVS, and DVS128 Gesture demonstrate that ParaRevSNN matches or exceeds the accuracy of standard RevSNNs, while reducing training time by up to 35.2\% and inference time to 18.15\%, making it well-suited for deployment in resource-constrained scenarios.
title ParaRevSNN: A Parallel Reversible Spiking Neural Network for Efficient Training and Inference
topic Computer Vision and Pattern Recognition
68T10
I.4.6
url https://arxiv.org/abs/2508.01223