Text-to-Video Retrieval on MSRVTT 1kA Query-Shift (test)
26.9Recall@1 (OCR)HAT-VTR
Evaluation Results
| Method | Links | ||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| HAT-VTRBackbone=X-Pool, Adaptation=TTA2026.02 | 26.9 | 14.8 | 14.4 | 18.5 | 14.7 | 44.8 | 43.8 | 44.4 | 46.6 | 47.3 | 42.6 | 48.7 | 46.1 | 46.1 | 48.1 | 36.5 | |
| TCRBackbone=X-Pool, Adaptation=TTA2026.02 | 25.7 | 12.8 | 13.2 | 16.4 | 12.2 | 43.6 | 43.9 | 41.5 | 45.2 | 46 | 42.7 | 46.1 | 44.5 | 44.8 | 46.8 | 35 | |
| SARBackbone=X-Pool, Adaptation=TTA2026.02 | 25.6 | 12.7 | 13.2 | 17.1 | 12.2 | 43.4 | 43.7 | 42.3 | 45.3 | 46.2 | 43.3 | 46.7 | 45.2 | 44.7 | 47 | 35.2 | |
| TentBackbone=X-Pool, Adaptation=TTA2026.02 | 25.5 | 12.4 | 13.1 | 17 | 12.3 | 43.2 | 43.8 | 42.2 | 45.2 | 46.2 | 42.9 | 46.7 | 44.9 | 44.6 | 46.7 | 35.1 | |
| EATABackbone=X-Pool, Adaptation=TTA2026.02 | 25.2 | 12.4 | 13.2 | 17.1 | 12.2 | 42.8 | 44.1 | 42.1 | 45.3 | 45.4 | 42.1 | 45.7 | 44.9 | 45 | 46.2 | 34.9 | |
| READBackbone=X-Pool, Adaptation=TTA2026.02 | 25.1 | 12.4 | 13.3 | 17 | 12 | 43.1 | 43.9 | 42.2 | 45.7 | 46.3 | 43.1 | 46.6 | 45.5 | 45.1 | 47.2 | 35.2 | |
| XpoolBackbone=X-Pool, Adaptation=None (Base Model)2026.02 | 25 | 12.5 | 13.2 | 16.9 | 12.1 | 43.4 | 44.1 | 42.3 | 45.3 | 46.2 | 43.1 | 46.6 | 45.3 | 44.9 | 47 | 35.2 | |
| HAT-VTRBackbone=CLIP4Clip, Adaptation=TTA2026.02 | 24.5 | 13.8 | 14.1 | 16.7 | 12.8 | 40.7 | 41.6 | 41.2 | 42.9 | 42.9 | 39.8 | 43.6 | 42.3 | 42.4 | 43.7 | 33.5 | |
| TCRBackbone=CLIP4Clip, Adaptation=TTA2026.02 | 21.7 | 12.8 | 13 | 14.9 | 11 | 39.1 | 39.4 | 39.4 | 39.8 | 39.8 | 37.8 | 40.9 | 39.8 | 40.1 | 41.7 | 31.4 | |
| EATABackbone=CLIP4Clip, Adaptation=TTA2026.02 | 21.6 | 12.5 | 12.2 | 15.1 | 10.5 | 38.6 | 39.3 | 39.7 | 40.2 | 39.8 | 38.4 | 41.2 | 39.6 | 40.2 | 41.5 | 31.4 | |
| CLIP4ClipBackbone=CLIP4Clip, Adaptation=None (Base Model)2026.02 | 21.5 | 12.8 | 12.2 | 15.7 | 11.2 | 38.9 | 39 | 39.7 | 39.5 | 39.7 | 38.5 | 40.9 | 39.7 | 40.6 | 41.8 | 31.4 | |
| TentBackbone=CLIP4Clip, Adaptation=TTA2026.02 | 21.5 | 12.6 | 11.9 | 15.9 | 11.1 | 38.6 | 39 | 39.9 | 39.7 | 39.5 | 38.3 | 41.4 | 40.1 | 41 | 41.6 | 31.5 | |
| READBackbone=CLIP4Clip, Adaptation=TTA2026.02 | 21.5 | 12.9 | 12.4 | 15.8 | 11 | 38.8 | 39.1 | 39.7 | 39.2 | 39.5 | 38.6 | 41 | 39.9 | 40.4 | 41.6 | 31.4 | |
| SARBackbone=CLIP4Clip, Adaptation=TTA2026.02 | 21.4 | 12.8 | 12.1 | 15.7 | 11.1 | 38.6 | 38.9 | 39.9 | 39.8 | 39.5 | 38.3 | 41.3 | 40.2 | 40 | 41.7 | 31.4 |