Cross-modal Retrieval on MOVi N=128 (test)
86.2R@1Omnivorous ViT-B/14
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| Omnivorous ViT-B/14feature type=gap2026.02 | 86.2 | 96.5 | 90.9 | 1 | |
| Omnivorous ViT-B/14feature type=tok2026.02 | 76.6 | 92.7 | 83.4 | 1 | |
| DINOv2 ViT-B/14feature type=tok2026.02 | 18.2 | 34.5 | 27.2 | 16.8 | |
| DINOv2 ViT-B/14feature type=gap2026.02 | 15.5 | 33.1 | 25.2 | 19.3 |