Visual Entity Recognition on EVQA
47.4Recall@1Google Lens
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| Google Lens2026.03 | 47.4 | 62.5 | 64.7 | 65.2 | |
| Echosight2026.03 | 36.5 | 47.9 | 48.8 | 48.8 | |
| WikiCLIP-LScale=Large2026.03 | 31.9 | 53.3 | 61.2 | 69 | |
| WikiCLIP-SScale=Small2026.03 | 30.7 | 53.6 | 62.5 | 69.1 | |
| CLIP I-IRetrieval Mode=Image-to-Image2026.03 | 13.3 | 31.3 | 41 | 48.8 | |
| CLIP I-TRetrieval Mode=Image-to-Text2026.03 | 3.3 | 7.7 | 12.1 | 16.5 |