Work

High-performance image classification via spiking vision transformer and an improved AA-LRSA attention mechanism