VideoMAE: Masked Autoencoders are Data-Efficient Learners for Self-Supervised Video Pre-Training
Paper • 2203.12602 • Published • 3
VideoMAE
VideoMAE Pre-trained Models
-
-
MCG-NJU/videomae-base
Video Classification • 94.2M • Updated • 60.1k • 52 -
MCG-NJU/videomae-base-finetuned-kinetics
Video Classification • 86.5M • Updated • 81.5k • 48 -
MCG-NJU/videomae-base-finetuned-ssv2
Video Classification • Updated • 2.64k • 7 -
MCG-NJU/videomae-large
Video Classification • 0.3B • Updated • 12.6k • 37 -
MCG-NJU/videomae-large-finetuned-kinetics
Video Classification • 0.3B • Updated • 6.42k • 14 -
MCG-NJU/videomae-small-finetuned-kinetics
Video Classification • Updated • 4.35k • 1 -
MCG-NJU/videomae-small-finetuned-ssv2
Video Classification • Updated • 401 -
MCG-NJU/videomae-huge-finetuned-kinetics
Video Classification • Updated • 344 • 4