Saved in:
Bibliographic Details
Main Authors: Zhang, Yuhao, Zhang, Chengjun, Wu, Di, Yang, Jie, Sawan, Mohamad
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2509.17461
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866911168239501312
author Zhang, Yuhao
Zhang, Chengjun
Wu, Di
Yang, Jie
Sawan, Mohamad
author_facet Zhang, Yuhao
Zhang, Chengjun
Wu, Di
Yang, Jie
Sawan, Mohamad
contents Spike-based transformer is a novel architecture aiming to enhance the performance of spiking neural networks while mitigating the energy overhead inherent to transformers. However, methods for generating these models suffer from critical limitations: excessive training costs introduced by direct training methods, or unavoidably hardware-unfriendly operations in existing conversion methods. In this paper, we propose CSDformer, a novel conversion method for fully spike-driven transformers. We tailor a conversion-oriented transformer-based architecture and propose a new function NReLU to replace softmax in self-attention. Subsequently, this model is quantized and trained, and converted into a fully spike-driven model with temporal decomposition technique. Also, we propose delayed Integrate-andFire neurons to reduce conversion errors and improve the performance of spiking models. We evaluate CSDformer on ImageNet, CIFAR-10 and CIFAR-100 datasets and achieve 76.36% top-1 accuracy under 7 time-steps on ImageNet, demonstrating superiority over state-of-the-art models. Furthermore, CSDformer eliminates the need for training SNNs, thereby reducing training costs (reducing computational resource by 75% and accelerating training speed by 2-3$\times$). To the best of our knowledge, this is the first fully spike-driven transformer-based model developed via conversion method, achieving high performance under ultra-low latency, while dramatically reducing both computational complexity and training overhead.
format Preprint
id arxiv_https___arxiv_org_abs_2509_17461
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle CSDformer: A Conversion Method for Fully Spike-Driven Transformer
Zhang, Yuhao
Zhang, Chengjun
Wu, Di
Yang, Jie
Sawan, Mohamad
Computer Vision and Pattern Recognition
Spike-based transformer is a novel architecture aiming to enhance the performance of spiking neural networks while mitigating the energy overhead inherent to transformers. However, methods for generating these models suffer from critical limitations: excessive training costs introduced by direct training methods, or unavoidably hardware-unfriendly operations in existing conversion methods. In this paper, we propose CSDformer, a novel conversion method for fully spike-driven transformers. We tailor a conversion-oriented transformer-based architecture and propose a new function NReLU to replace softmax in self-attention. Subsequently, this model is quantized and trained, and converted into a fully spike-driven model with temporal decomposition technique. Also, we propose delayed Integrate-andFire neurons to reduce conversion errors and improve the performance of spiking models. We evaluate CSDformer on ImageNet, CIFAR-10 and CIFAR-100 datasets and achieve 76.36% top-1 accuracy under 7 time-steps on ImageNet, demonstrating superiority over state-of-the-art models. Furthermore, CSDformer eliminates the need for training SNNs, thereby reducing training costs (reducing computational resource by 75% and accelerating training speed by 2-3$\times$). To the best of our knowledge, this is the first fully spike-driven transformer-based model developed via conversion method, achieving high performance under ultra-low latency, while dramatically reducing both computational complexity and training overhead.
title CSDformer: A Conversion Method for Fully Spike-Driven Transformer
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2509.17461