Video-to-text retrieval on MSRVTT severity degree 1 (1kA)
44.9Score (Gaussian)HAT-VTR
Evaluation Results
| Method | Links | |||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| HAT-VTRBase Model=SAR, Adaptation=HAT-VTR2026.02 | 44.9 | 40.4 | 45.7 | 42.3 | 44.6 | 47.9 | 45.1 | 41.6 | 44 | 38.2 | 42.4 | 42.6 | 43.3 | |
| EATABase Model=SAR, Adaptation=EATA2026.02 | 43.1 | 35.3 | 43.5 | 36.8 | 42.5 | 44.8 | 42.5 | 38.5 | 39.7 | 34.9 | 39.6 | 40.5 | 40.1 | |
| TCRBase Model=SAR, Adaptation=TCR2026.02 | 41.7 | 33.8 | 42.5 | 35.6 | 42.8 | 44.6 | 42.5 | 37.8 | 40.2 | 34.4 | 40.1 | 39.5 | 39.6 | |
| HAT-VTRBase Model=CLIP4Clip, Adaptation=HAT-VTR2026.02 | 41.1 | 38.7 | 41.8 | 38.9 | 40.7 | 42.5 | 41.9 | 37.4 | 39.3 | 35.6 | 40.6 | 38.8 | 39.8 | |
| TentBase Model=Xpool, Adaptation=Tent2026.02 | 41.1 | 32 | 41.7 | 33.8 | 44.2 | 45.8 | 41.9 | 38.3 | 40 | 33.4 | 40.4 | 40.3 | 39.4 | |
| SARBase Model=SAR, Adaptation=None2026.02 | 41.1 | 32 | 41.7 | 33.2 | 44.1 | 45.5 | 42.2 | 37.7 | 40.3 | 33.2 | 40.4 | 40.2 | 39.3 | |
| READBase Model=Xpool, Adaptation=READ2026.02 | 40.7 | 30 | 41 | 31.8 | 43 | 44.7 | 41.7 | 35.5 | 38.9 | 29.8 | 39.8 | 39.1 | 38 | |
| XpoolBase Model=Xpool, Adaptation=None2026.02 | 40.5 | 30.6 | 41.6 | 32.1 | 42.9 | 45 | 41.7 | 36.5 | 39.1 | 31.3 | 40.4 | 39.4 | 38.4 | |
| EATABase Model=CLIP4Clip, Adaptation=EATA2026.02 | 39 | 31.1 | 40.2 | 33.2 | 39.5 | 42.7 | 39.4 | 34.9 | 36.2 | 31.5 | 38.8 | 39.3 | 37.2 | |
| TentBase Model=CLIP4Clip, Adaptation=Tent2026.02 | 38.8 | 27.1 | 39.4 | 30.3 | 39.6 | 42 | 39.7 | 33.9 | 35.1 | 30.4 | 37.9 | 38.8 | 36.1 | |
| TCRBase Model=CLIP4Clip, Adaptation=TCR2026.02 | 38.8 | 33.6 | 38.7 | 34.2 | 38.4 | 40.9 | 37.6 | 33.1 | 36.6 | 31.3 | 36.6 | 38.6 | 36.5 | |
| SARBase Model=CLIP4Clip, Adaptation=SAR2026.02 | 38.5 | 27.9 | 39.5 | 30.6 | 39.8 | 42.2 | 39.7 | 33.9 | 35 | 30.5 | 37.8 | 38.8 | 36.2 | |
| CLIP4ClipBase Model=CLIP4Clip, Adaptation=None2026.02 | 38.3 | 26.8 | 38.3 | 29.6 | 39 | 42 | 39.2 | 32.3 | 34.9 | 28.6 | 37.6 | 39 | 35.5 | |
| READBase Model=CLIP4Clip, Adaptation=READ2026.02 | 37.3 | 26.6 | 37.8 | 29.4 | 38.7 | 41.7 | 39 | 31.3 | 34.1 | 26.5 | 36.9 | 38.8 | 34.8 |