Video Retrieval (clip-caption) on YouCookII (evaluation)
24.6Recall@1MMFT
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| MMFTbackbone=S3D-G2022.12 | 24.6 | 48.3 | 60.4 | |
| TempCLRbackbone=S3D-G2022.12 | 23.3 | 51 | 64.5 | |
| VideoCLIPbackbone=S3D-G2022.12 | 22.7 | 50.4 | 63.1 | |
| MCNbackbone=R152+RX1012022.12 | 18.1 | 35.5 | 45.2 | |
| MIL-NCEbackbone=S3D-G2022.12 | 15.1 | 38 | 51.2 | |
| MMVbackbone=TSM-50x22022.12 | 11.7 | 33.4 | 45.4 | |
| MIL-NCEbackbone=I3D-G2022.12 | 11.4 | 30.6 | 42 | |
| ActBERTbackbone=R101+Res3D2022.12 | 9.6 | 26.7 | 38 | |
| MIL-NCE*backbone=R152+RX1012022.12 | 8.1 | 23.3 | 32.3 | |
| HT100Mbackbone=S3D-G2022.12 | 6.1 | 17.3 | 24.8 | |
| Randombackbone=-2022.12 | 0 | 0.2 | 0.3 |