Image-to-Text Retrieval on SKETCHY
68.22R@1StructXLIP
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| StructXLIPfinetuning_mode=Full-parameter2026.02 | 68.22 | 90.67 | 95.68 | |
| GOALfinetuning_mode=Full-parameter2026.02 | 62.44 | 87.82 | 92.31 | |
| Long-CLIPfinetuning_mode=Full-parameter2026.02 | 52.76 | 80.31 | 88.08 | |
| SmartCLIPfinetuning_mode=Full-parameter2026.02 | 51.3 | 80.83 | 94.04 | |
| FineLIPfinetuning_mode=Full-parameter2026.02 | 40.33 | 72.11 | 82.38 |