Publication | Closed Access

STAR++: Rethinking spatio-temporal cross attention transformer for video action recognition