Video-to-Text Retrieval on MSRVTT (R@1, R@5, R@10)
40.4Recall@1HAT-VTR
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| HAT-VTRBase Model=Xpool, Adaptation Method=HAT-VTR2026.02 | 40.4 | 65.6 | 77.3 | |
| HAT-VTRBase Model=CLIP4Clip, Adaptation Method=HAT-VTR2026.02 | 37.5 | 64.4 | 74.1 | |
| TCRBase Model=Xpool, Adaptation Method=TCR2026.02 | 36.8 | 62.3 | 74.3 | |
| EATABase Model=Xpool, Adaptation Method=EATA2026.02 | 36.6 | 62.9 | 75.2 | |
| HAT-VTRBackbone=CLIP4Clip2026.02 | 36.6 | 61.7 | 72 | |
| SARBase Model=Xpool, Adaptation Method=SAR2026.02 | 35.8 | 62.6 | 73.7 | |
| TentBase Model=Xpool, Adaptation Method=Tent2026.02 | 35.7 | 62.8 | 73.6 | |
| XpoolBase Model=Xpool, Adaptation Method=None2026.02 | 35.5 | 61.8 | 73.2 | |
| HAT-VTRBackbone=Xpool2026.02 | 35.3 | 60.2 | 69.6 | |
| READBase Model=Xpool, Adaptation Method=READ2026.02 | 35.2 | 61.2 | 73.1 | |
| EATABase Model=CLIP4Clip, Adaptation Method=EATA2026.02 | 33.7 | 60.1 | 71.4 | |
| TCRBase Model=CLIP4Clip, Adaptation Method=TCR2026.02 | 33.1 | 57.8 | 69.3 | |
| TCRBackbone=CLIP4Clip2026.02 | 32.6 | 58.6 | 68.6 | |
| TentBase Model=CLIP4Clip, Adaptation Method=Tent2026.02 | 32.5 | 58.5 | 69.3 | |
| SARBase Model=CLIP4Clip, Adaptation Method=SAR2026.02 | 32.2 | 58.5 | 69.6 | |
| CLIP4ClipBase Model=CLIP4Clip, Adaptation Method=None2026.02 | 32 | 57.9 | 68.5 | |
| READBase Model=CLIP4Clip, Adaptation Method=READ2026.02 | 31.7 | 57.2 | 68.4 | |
| EATABackbone=CLIP4Clip2026.02 | 31.6 | 58.5 | 68.4 | |
| TentBackbone=CLIP4Clip2026.02 | 30.9 | 56.8 | 67.3 | |
| SARBackbone=CLIP4Clip2026.02 | 30.8 | 56.8 | 67.3 | |
| CLIP4ClipBackbone=CLIP4Clip2026.02 | 30.3 | 57.1 | 66.9 | |
| READBackbone=Xpool2026.02 | 30.2 | 54.1 | 65.1 | |
| EATABackbone=Xpool2026.02 | 30.2 | 54.5 | 65.8 | |
| TCRBackbone=Xpool2026.02 | 30.1 | 54.5 | 65.6 | |
| SARBackbone=Xpool2026.02 | 29.9 | 53.9 | 65.4 | |
| XpoolBackbone=Xpool2026.02 | 29.8 | 54 | 65.2 | |
| TentBackbone=Xpool2026.02 | 29.7 | 53.9 | 65.3 | |
| READBackbone=CLIP4Clip2026.02 | 29.1 | 56.4 | 66.6 |