Text-to-Image Retrieval on DCI (test)
59.73R@1FAST-GOAL
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| FAST-GOALBackbone=ViT-L/142026.05 | 59.73 | 77.23 | 88.21 | 91.45 | |
| GOALBackbone=ViT-L/142026.05 | 54.66 | 74.57 | 86.29 | 90.54 | |
| Long-CLIPBackbone=ViT-L/142026.05 | 54.31 | 72.53 | 85.04 | 89.06 | |
| EVA-CLIPBackbone=ViT-L/142026.05 | 45.88 | 64.96 | 79.56 | 84.87 | |
| FineCLIPBackbone=ViT-L/142026.05 | 42.12 | 62.01 | 78.11 | 83.94 | |
| CLIPBackbone=ViT-L/142026.05 | 31.18 | 50.34 | 68.84 | 76.15 |