| Dataset Name | SOTA Method | Metric | Trend | ||
|---|---|---|---|---|---|
| Kinetics-400 | VideoMAE V2 | Top-1 Acc90 | 184 | 2d ago | |
| UCF101 | Ours ViT-L | Top-1 Acc96.9 | 153 | 2d ago | |
| Kinetics-400 (val) | FTP-UniFormerV2-L/14 | Top-1 Acc94.3 | 151 | 3d ago | |
| HMDB-51 (3 splits) | SCT-L | Accuracy84.6 | 116 | 3d ago | |
| HMDB51 | OV-Encoder (Codec) | Top-1 Accuracy85.3 | 103 | 2d ago | |
| HMDB51 (test) | SMART | Accuracy84.3 | 73 | 3d ago | |
| HMDB51 (avg over all splits) | Our (MViTv2-S based) | Top-1 Acc83.39 | 56 | 3d ago | |
| Kinetics-600 | EVA | Top-1 Acc89.8 | 48 | 3d ago | |
| UCF101 (test) | CVRL | Top-1 Acc92.1 | 46 | 3d ago | |
| Kinetics 400 (test) | CAST | Top-1 Accuracy85.3 | 44 | 3d ago | |
| UCF101 avg over all splits | BraVe | Top-1 Accuracy96.9 | 42 | 3d ago | |
| HMDB-51 | S3D-G | Mean Class Accuracy75.9 | 37 | 3d ago | |
| UCF101 5-way 5-shot | DIST | Accuracy99.2 | 28 | 3d ago | |
| HMDB51 5-way 5-shot | DIST | Accuracy88.7 | 28 | 3d ago | |
| Kinetics-400 v1 (test) | MVFNet_En | Top-1 Accuracy79.1 | 26 | 3d ago | |
| Epic-Kitchens 100 (test) | ORViT MF-HR-Light | Top-1 Action Accuracy46.1 | 24 | 3d ago | |
| Kinetics | TRX | Accuracy85.9 | 23 | 3d ago | |
| Kinetics-700 | EVA | Top-1 Acc82.9 | 20 | 3d ago | |
| CharadesEgo | DINOv3 | Top-1 Accuracy14 | 16 | 3d ago | |
| Perception Test | OV-Encoder (Codec) | Top-1 Accuracy60.9 | 16 | 3d ago | |
| SS v2 | TC-CLIP | Base Score19.6 | 15 | 3d ago | |
| HMDB51 | MCN | Accuracy54.8 | 13 | 3d ago | |
| UCF101 -> HMDB51 | Top-1 Accuracy44.25 | 13 | 3d ago | ||
| Kinetics-600 5 (test) | Video-FocalNet-B | Top-1 Accuracy86.7 | 13 | 3d ago | |
| EPIC-KITCHENS-100 | SMILE | Top-1 Acc63.3 | 12 | 3d ago |