| Dataset Name | SOTA Method | Metric | Trend | ||
|---|---|---|---|---|---|
| nocaps | OVI (T→I) | mAP43.7 | 12 | 26d ago | |
| Flickr30k | IsoCLIP | mAP60.8 | 12 | 26d ago | |
| COCO | OVI (T→I) | mAP31.9 | 12 | 26d ago | |
| Crisscrossed Captions (CxC) | ALIGN-L2 | R@161.8 | 10 | 1mo ago | |
| COCO-QLTI | Ours-stage3 | R@1083.4 | 6 | 1mo ago | |
| NLP Retrieval Benchmarks standard (test) | SeMoBridge-T | IMDB Retrieval Score57.42 | 4 | 8d ago |