Video-to-Text Retrieval on MSR-VTT 1K (val)
0.562R@1Weighted Token-wise Interaction (WTI)
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| Weighted Token-wise Interaction (WTI)Backbone=ViT-B/16, Normalization=QB-Norm2022.03 | 0.562 | 0.799 | 0.874 | 1 | |
| Weighted Token-wise Interaction (WTI)Backbone=ViT-B/162022.03 | 0.489 | 0.763 | 0.854 | 2 | |
| Weighted Token-wise Interaction (WTI)2022.03 | 0.453 | 0.739 | 0.833 | 2 | |
| CLIP4Clip2022.03 | 0.427 | 0.709 | 0.806 | 2 | |
| MDMMT2022.03 | 0.27 | 0.575 | 0.697 | 3.7 | |
| SUPPORT2022.03 | 0.266 | 0.551 | 0.675 | 3 |