Efficient Video Action Detection with Token Dropout and Context Refinement.

Lei Chen, Zhan Tong, Yibing Song, Gangshan Wu, Limin Wang

Browse the full ICCV paper archive.