Text-to-Video Retrieval on DiDeMo (val)
53.9R@1SINGULARITY
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| SINGULARITY#PT=17M, #Train Frame=12022.06 | 53.9 | 79.4 | 86.9 | |
| SINGULARITY-temporal#PT=17M, #Train Frame=42022.06 | 53.1 | 79.9 | 88.1 | |
| SINGULARITY-temporal#PT=5M, #Train Frame=42022.06 | 49.2 | 77.5 | 85.4 | |
| SINGULARITY#PT=5M, #Train Frame=12022.06 | 47.4 | 75.2 | 84 | |
| CLIP4Clip#PT=400M, #Train Frame=12/64/642022.06 | 42.8 | 68.5 | 79.2 | |
| AlignPrompt#PT=5M, #Train Frame=82022.06 | 35.9 | 67.5 | 78.8 | |
| Frozen#PT=5M, #Train Frame=42022.06 | 31 | 59.8 | 72.4 | |
| ClipBERT#PT=0.2M, #Train Frame=16/16/82022.06 | 20.4 | 48 | 60.8 |