Cross-Modal Learning with 3D Deformable Attention for Action Recognition

ViT
3DPC
Re-assign community