Video-to-Text Retrieval on MSRVTT (R@1, R@5)
0.354Recall@1HAT-VTR
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| HAT-VTRZero-shot adaptation=true, QGS=true2026.02 | 0.354 | 0.614 | |
| TCRZero-shot adaptation=true, QGS=true2026.02 | 0.279 | 0.548 | |
| EATAZero-shot adaptation=true, QGS=true2026.02 | 0.272 | 0.535 | |
| TentZero-shot adaptation=true, QGS=true2026.02 | 0.268 | 0.522 | |
| SARZero-shot adaptation=true, QGS=true2026.02 | 0.267 | 0.525 | |
| CLIPZero-shot adaptation=true, QGS=true2026.02 | 0.265 | 0.518 | |
| READZero-shot adaptation=true, QGS=true2026.02 | 0.257 | 0.493 |