| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Action Recognition | Assembly101 | Action Accuracy47.51 | 18 | |
| Action Recognition | Assembly101 (test) | Top-1 Accuracy62.7 | 13 | |
| Procedural mistake detection | Assembly101-O | F1 Score59.6 | 13 | |
| Temporal Action Segmentation | Assembly101 view C10119 v4 (val) | Accuracy43 | 12 | |
| Action Anticipation | Assembly101 (val) | Recall@5 (Action, Overall)12.07 | 8 | |
| Action Segmentation | Assembly101 (test) | F1@1033.9 | 5 | |
| Exocentric-to-Egocentric Translation | Assembly101 (unseen actions) | FID50.232 | 4 | |
| Temporal Action Segmentation | Assembly101 | F1 Score @ 10%19.6 | 3 | |
| Video Retrieval | Assembly101 (test) | mAP25.3 | 3 | |
| Next-primitive prediction | Assembly101 | Top-1 Accuracy54.5 | 3 | |
| Action Recognition | Assembly101-OV (HM) | Verb Top-1 Acc50.5 | 3 | |
| Action Recognition | Assembly101 OV (Novel) | Verb Top-1 Acc45.1 | 3 | |
| Action Recognition | Assembly101-OV (Closed) | Verb Top-1 Acc57.6 | 3 |