SMAUG: Sparse Masked Autoencoder for Efficient Video-Language Pre-training.

Yuanze Lin, Chen Wei, Huiyu Wang, Alan L. Yuille, Cihang Xie

Browse the full ICCV paper archive.