Enhancing Transformer for Video Understanding Using Gated Multi-Level Attention and Temporal Adversarial Training

ViT
Re-assign community