| action-recognition-in-videos-on-ucf101 | VideoMAE V2-g | #1 | 3-fold Accuracy: 99.6 |
| self-supervised-action-recognition-on-ucf101 | VideoMAE V2-g | #1 | 3-fold Accuracy: 99.6 |
| spatio-temporal-action-localization-on-ava-kinetics | VideoMAE V2-g | #1 | val mAP: 42.6 |
| action-recognition-in-videos-on-hmdb-51 | VideoMAE V2-g | #2 | Average accuracy of 3 splits: 88.7 |
| action-recognition-in-videos-on-something-something-v1 | VideoMAE V2-g | #2 | Top 1 Accuracy: 68.7Top 5 Accuracy: 91.9 |
| action-recognition-on-ava-v2-2 | VideoMAE V2-g | #3 | mAP: 42.6 |
| action-recognition-in-videos-on-something-something-v2 | VideoMAE V2-g | #5 | Top-1 Accuracy: 77.0Top-5 Accuracy: 95.9Parameters: 1013… |
| temporal-action-localization-on-fineaction | VideoMAE V2-g | #5 | mAP: 18.24mAP IOU@0.5: 29.07mAP IOU@0.75: 17.66mAP IOU@0.95: 5.07 |
| temporal-action-localization-on-thumos14 | ActionFormer (VideoMAE V2-g features) | #8 | Avg mAP (0.3:0.7): 69.6mAP IOU@0.3: 84.0mAP IOU@0.4: 79.6… |
| action-classification-on-kinetics-400 | VideoMAE V2-g (64x266x266) | #9 | Acc@1: 90.0Acc@5: 98.4 |
| action-classification-on-kinetics-400 | VideoMAE V2-g | #22 | Acc@1: 88.5Acc@5: 98.1 |
| action-classification-on-kinetics-600 | VideoMAE V2-g (64x266x266) | #10 | Top-1 Accuracy: 89.9Top-5 Accuracy: 98.5 |
| action-classification-on-kinetics-600 | VideoMAE V2-g | #14 | Top-1 Accuracy: 88.8Top-5 Accuracy: 98.2 |