| Paper | Code | Avg mAP (0.1-0.5) | mAP IOU@0.1 | mAP IOU@0.2 | mAP IOU@0.3 | mAP IOU@0.4 | mAP IOU@0.5 | ModelName | ReleaseDate |
|---|---|---|---|---|---|---|---|---|---|
| End-to-End Temporal Action Detection with 1B Parameters Across 1000 Frames | ✓ Link | 29.3 | 33.1 | 32.2 | 30.4 | 27.5 | 23.1 | AdaTAD (verb, VideoMAE-L) | 2023-11-28 |
| TriDet: Temporal Action Detection with Relative Boundary Modeling | ✓ Link | 25.4 | 28.6 | 27.4 | 26.1 | 24.2 | 20.8 | TriDet (verb) | 2023-03-13 |
| TemporalMaxer: Maximize Temporal Context with only Max Pooling for Temporal Action Localization | ✓ Link | 24.5 | 27.8 | 26.6 | 25.3 | 23.1 | 19.9 | TemporalMaxer (verb) | 2023-03-16 |
| ActionFormer: Localizing Moments of Actions with Transformers | ✓ Link | 23.5 | 26.6 | 25.4 | 24.2 | 22.3 | 19.1 | ActionFormer (verb) | 2022-02-16 |
| G-TAD: Sub-Graph Localization for Temporal Action Detection | ✓ Link | 9.4 | 12.1 | 11.0 | 9.4 | 8.1 | 6.5 | G-TAD (verb) | 2019-11-26 |
| BMN: Boundary-Matching Network for Temporal Action Proposal Generation | ✓ Link | 8.4 | 10.8 | 9.8 | 8.4 | 7.1 | 5.6 | BMN (verb) | 2019-07-23 |