ST2ST: Self-Supervised Test-time Adaptation for Video Action Recognition.

Masud An Nur Islam Fahim, Mohammed Innat, Jani Boutellier

Browse the full CVPR paper archive.