Cross-modal Retrieval on MSCOCO V-T Zero-Shot
1.3R@1MICU
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| MICUTraining Dataset=VGGSound-AVEL 40K, Evaluation Protocol=Zero-shot2026.05 | 1.3 | 5 | 8.8 | 7.82 | |
| CoDAARTraining Dataset=VGGSound-AVEL 90K, Evaluation Protocol=Zero-shot2026.05 | 1.3 | 6.1 | 10.4 | 9.45 | |
| MICUTraining Dataset=VGGSound-AVEL 90K, Evaluation Protocol=Zero-shot2026.05 | 0.9 | 4.9 | 9.5 | 7.82 | |
| DCIDTraining Dataset=VGGSound-AVEL 40K, Evaluation Protocol=Zero-shot2026.05 | 0.8 | 5 | 8.3 | 7.49 | |
| CoDAARTraining Dataset=VGGSound-AVEL 40K, Evaluation Protocol=Zero-shot2026.05 | 0.8 | 4.4 | 9.5 | 9.45 | |
| DCIDTraining Dataset=VGGSound-AVEL 90K, Evaluation Protocol=Zero-shot2026.05 | 0.8 | 4.8 | 8.2 | 7.49 |