Video-to-Text Retrieval on MSR-VTT (test)
50.3R@1CAMoE
Evaluation results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| CAMoE2021.09 | 50.3 | 74.6 | — | — | — | — | |
| CoCaZero-shot=true, subset=true2022.05 | 49.9 | 73.4 | 81.4 | — | — | — | |
| Socratic ModelsZero-shot=true, subset=true2022.05 | 46.9 | 73.5 | 81.3 | — | — | — | |
| Socratic ModelsZero-shot=true, subset=false2022.05 | 44.7 | 71.2 | 80 | — | — | — | |
| CLIPZero-shot=true, subset=true2022.05 | 43.3 | 73.3 | 81.8 | — | — | — | |
| CLIPZero-shot=true, subset=false2022.05 | 40.3 | 69.7 | 79.2 | — | — | — | |
| Dual EncodingSource=Proposed, Visual Feature=ResNeXt-ResNet2020.09 | 22.5 | 47.1 | 58.9 | 7 | 10.5 | 211.7 | |
| HGRSource=Re-trained, Visual Feature=ResNeXt-ResNet2020.09 | 18.7 | 44.3 | 57.6 | 7 | 9.9 | 204.4 | |
| W2VV++Source=Re-trained, Visual Feature=ResNeXt-ResNet2020.09 | 17.5 | 40.2 | 52.5 | 9 | 8.5 | 191.4 | |
| W2VVSource=Re-trained, Visual Feature=ResNeXt-ResNet2020.09 | 17 | 37.9 | 49.1 | 11 | 7.6 | 117.9 | |
| CESource=Original Paper2020.09 | 15.6 | 40.9 | 55.2 | 8.3 | — | 191.9 | |
| VSE++Source=Re-trained, Visual Feature=ResNeXt-ResNet2020.09 | 15.6 | 36.6 | 48.6 | 11 | 7.4 | 167.9 | |
| TCESource=Re-trained, Visual Feature=ResNeXt-ResNet2020.09 | 15.1 | 36.8 | 50.2 | 10 | 8 | 177.3 | |
| HGRSource=Original Paper2020.09 | 15 | 36.7 | 48.8 | 11 | — | 172.4 | |
| MEESource=Re-trained, Visual Feature=ResNeXt-ResNet2020.09 | 13.4 | 32 | 44 | 14 | 6.6 | 148 | |
| Mithun et al.Source=Original Paper2020.09 | 12.5 | 32.1 | 42.4 | 16 | — | 144.6 | |
| CESource=Re-trained, Visual Feature=ResNeXt-ResNet2020.09 | 11 | 31.9 | 46.1 | 13 | 6.8 | 155.1 |