BViT: Broad Attention based Vision Transformer

Nannan Li, Yaran Chen, Weifan Li +2

Recent works have demonstrated that transformer can achieve promising performance in computer vision, by exploiting the relationship among image patches with self-attention. While…