Content-aware Token Sharing for Efficient Semantic Segmentation with Vision Transformers

ViT
Re-assign community