Text-Video Retrieval on MSR-VTT (split1)
28.4R@1TACo
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| TACoEvaluation Protocol=Finetuned, Video Backbone=R-152+S3D-HM2021.08 | 28.4 | 57.8 | 71.2 | 4 | |
| TACoLang.=BERT, Video=R-152+S3D-HM2021.08 | 26.7 | 54.5 | 68.2 | 4 | |
| MMTEvaluation Protocol=Finetuned, Video Backbone=Collaborative Experts2021.08 | 26.6 | 57.1 | 69.6 | 4 | |
| MMTLang.=BERT, Video=Collaborative Experts2021.08 | 24.6 | 54 | 67.1 | 4 | |
| CELang.=GPT, Video=Collaborative Experts2021.08 | 20.9 | 48.8 | 62.4 | 6 |