Existing FER methods typically report overall accuracy on highly imbalanced test sets but exhibit low performance in terms of the mean accuracy across all expression classes.

Conventional wisdom suggests that pre-training Vision Transformers (ViT) improves downstream performance by learning useful representations.

CLIP), have attracted widespread attention and adoption across various domains. Nonetheless, CLIP has been observed to be susceptible to adversarial examples.

The spike firing rate is defined as the proportion of non-zero elements in the spike tensor. In Table S1, we present the spike firing rates for all spiking tensors in spike-driven Transformer-8-512. SNNs are theoretically more energy efficient than counterpart ANNs. We employ two types of datasets: static image classification and neuromorphic classification. ImageNet-1K is the most typical static image dataset, which is widely used in the field of image classification.