| Dataset Name | SOTA Method | Metric | Trend | ||
|---|---|---|---|---|---|
| Image-Text-Audio | UnifiedIO2-L | CIDEr91.36 | 9 | 1mo ago | |
| Storyboard20K | DreamBooth | CIDEr41.55 | 7 | 4mo ago | |
| Chart2Text | DATAPROPHET | Score72 | 5 | 4mo ago | |
| one-to-one evaluation benchmarks | CLIP Score32.14 | 5 | 4mo ago | ||
| One-to-one evaluation benchmarks Image-to-Text | UnifiedIO2-L | CIDEr134.7 | 5 | 4mo ago | |
| MS-COCO 5% 2014 (test val) | GIT-large | CLIP Similarity (CLIP)25.5 | 4 | 1mo ago | |
| CUB | Ours (EBM) | CLIP I2T Score29 | 4 | 2mo ago |