Spiking Transformer with Spatial-Temporal Attention

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Lee, Donghyun, Li, Yuhang, Kim, Youngeun, Xiao, Shiting, Panda, Priyadarshini
Natura: Preprint
Pubblicazione: 2024
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866915180159434752
author Lee, Donghyun
Li, Yuhang
Kim, Youngeun
Xiao, Shiting
Panda, Priyadarshini
author_facet Lee, Donghyun
Li, Yuhang
Kim, Youngeun
Xiao, Shiting
Panda, Priyadarshini
contents Spike-based Transformer presents a compelling and energy-efficient alternative to traditional Artificial Neural Network (ANN)-based Transformers, achieving impressive results through sparse binary computations. However, existing spike-based transformers predominantly focus on spatial attention while neglecting crucial temporal dependencies inherent in spike-based processing, leading to suboptimal feature representation and limited performance. To address this limitation, we propose Spiking Transformer with Spatial-Temporal Attention (STAtten), a simple and straightforward architecture that efficiently integrates both spatial and temporal information in the self-attention mechanism. STAtten introduces a block-wise computation strategy that processes information in spatial-temporal chunks, enabling comprehensive feature capture while maintaining the same computational complexity as previous spatial-only approaches. Our method can be seamlessly integrated into existing spike-based transformers without architectural overhaul. Extensive experiments demonstrate that STAtten significantly improves the performance of existing spike-based transformers across both static and neuromorphic datasets, including CIFAR10/100, ImageNet, CIFAR10-DVS, and N-Caltech101. The code is available at https://github.com/Intelligent-Computing-Lab-Yale/STAtten
format Preprint
id arxiv_https___arxiv_org_abs_2409_19764
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Spiking Transformer with Spatial-Temporal Attention
Lee, Donghyun
Li, Yuhang
Kim, Youngeun
Xiao, Shiting
Panda, Priyadarshini
Neural and Evolutionary Computing
Spike-based Transformer presents a compelling and energy-efficient alternative to traditional Artificial Neural Network (ANN)-based Transformers, achieving impressive results through sparse binary computations. However, existing spike-based transformers predominantly focus on spatial attention while neglecting crucial temporal dependencies inherent in spike-based processing, leading to suboptimal feature representation and limited performance. To address this limitation, we propose Spiking Transformer with Spatial-Temporal Attention (STAtten), a simple and straightforward architecture that efficiently integrates both spatial and temporal information in the self-attention mechanism. STAtten introduces a block-wise computation strategy that processes information in spatial-temporal chunks, enabling comprehensive feature capture while maintaining the same computational complexity as previous spatial-only approaches. Our method can be seamlessly integrated into existing spike-based transformers without architectural overhaul. Extensive experiments demonstrate that STAtten significantly improves the performance of existing spike-based transformers across both static and neuromorphic datasets, including CIFAR10/100, ImageNet, CIFAR10-DVS, and N-Caltech101. The code is available at https://github.com/Intelligent-Computing-Lab-Yale/STAtten
title Spiking Transformer with Spatial-Temporal Attention
topic Neural and Evolutionary Computing
url https://arxiv.org/abs/2409.19764