标题： FSTA-SNN：基于频率的时空注意力模块用于脉冲神经网络 显示英文标题

标题： FSTA-SNN:Frequency-based Spatial-Temporal Attention Module for Spiking Neural Networks

摘要： 脉冲神经网络（SNNs）由于其固有的能效优势，正逐渐成为人工神经网络（ANNs）的有前景的替代方案。由于SNNs中脉冲生成的固有稀疏性，对中间输出脉冲的深入分析和优化往往被忽视。这种疏忽显著限制了SNNs的固有能效，并削弱了它们在时空特征提取中的优势，导致准确性不足和不必要的能量消耗。在本工作中，我们从时间和空间两个角度分析了SNNs的固有脉冲特性。在空间分析方面，我们发现浅层倾向于学习垂直变化，而深层逐渐学习特征的水平变化。关于时间分析，我们观察到不同时间步长之间的特征学习没有显著差异。这表明增加时间步长对特征学习的影响有限。基于这些分析所得的见解，我们提出了一种基于频率的空间-时间注意力（FSTA）模块，以增强SNNs中的特征学习。该模块旨在通过抑制冗余的脉冲特征来提高特征学习能力。实验结果表明，引入FSTA模块显著降低了SNNs的脉冲发放率，在多个数据集上表现出优于最先进基线的性能。我们的源代码可在 https://github.com/yukairong/FSTA-SNN 中获得。 显示英文摘要

摘要： Spiking Neural Networks (SNNs) are emerging as a promising alternative to Artificial Neural Networks (ANNs) due to their inherent energy efficiency.Owing to the inherent sparsity in spike generation within SNNs, the in-depth analysis and optimization of intermediate output spikes are often neglected.This oversight significantly restricts the inherent energy efficiency of SNNs and diminishes their advantages in spatiotemporal feature extraction, resulting in a lack of accuracy and unnecessary energy expenditure.In this work, we analyze the inherent spiking characteristics of SNNs from both temporal and spatial perspectives.In terms of spatial analysis, we find that shallow layers tend to focus on learning vertical variations, while deeper layers gradually learn horizontal variations of features.Regarding temporal analysis, we observe that there is not a significant difference in feature learning across different time steps.This suggests that increasing the time steps has limited effect on feature learning.Based on the insights derived from these analyses, we propose a Frequency-based Spatial-Temporal Attention (FSTA) module to enhance feature learning in SNNs.This module aims to improve the feature learning capabilities by suppressing redundant spike features.The experimental results indicate that the introduction of the FSTA module significantly reduces the spike firing rate of SNNs, demonstrating superior performance compared to state-of-the-art baselines across multiple datasets.Our source code is available in https://github.com/yukairong/FSTA-SNN.