Progressive Mask Distillation for Self-supervised Video Representation
Open paper
Benchmark
Model
Rank
Results
action-classification-on-kinetics-400
PMD 400e (ViT-L)
–
Acc@1: 86.3
action-recognition-in-videos-on-something-something-v2
PMD 400e (ViT-L)
–
Top-1 Accuracy: 76.4
Rank counts only results with a code link.