| Dataset Name | SOTA Method | Metric | Trend | ||
|---|---|---|---|---|---|
| Street View Text (SVT) | CA-FCN | Accuracy98.8 | 80 | 3mo ago | |
| IIIT, SVT, IC13, IC15, SVTP, CT | CCD-ViT-Small | IIIT Acc98 | 37 | 3mo ago | |
| Chinese text recognition benchmark | DTrOCR | Scene Acc87.4 | 33 | 3mo ago | |
| SVTP | MaskOCR (ViT-S) | Word Accuracy94.9 | 22 | 3mo ago | |
| IC15 | MaskOCR (ViT-S) | Word Accuracy90.2 | 22 | 3mo ago | |
| IIIT5K | MaskOCR (ViT-S) | Word Accuracy98 | 22 | 3mo ago | |
| SVT | MaskOCR (ViT-S) | Word Accuracy96.9 | 22 | 3mo ago | |
| IC13 | TrOCR | Word Accuracy98.3 | 22 | 3mo ago | |
| SROIE Task 2 (test) | DTrOCR | F1 Score98.37 | 19 | 3mo ago | |
| ICDAR 2003 | Proposed | Accuracy98.7 | 19 | 3mo ago | |
| ICDAR 2013 | Bai et al. 2018 | Accuracy94.4 | 15 | 3mo ago | |
| Unitail | ABINet | NED0.11 | 12 | 3mo ago | |
| IIIT 5k-word | Proposed | Accuracy97.1 | 11 | 3mo ago | |
| TextOCR (test) | TPS-ResNet-BiLSTM-Attn | Word Accuracy69.49 | 10 | 3mo ago | |
| IIIT (test) | Word Accuracy90.3 | 10 | 3mo ago | ||
| OmniDocBench v1.6 (Full) | MinerU2.5-Pro | Edit Distance1.9 | 9 | 1mo ago | |
| OmniDocBench v1.6 (Hard) | MinerU2.5-Pro | Edit Distance4.8 | 9 | 1mo ago | |
| OmniDocBench Base v1.6 | MinerU2.5-Pro | Edit Distance1.5 | 9 | 1mo ago | |
| BCTR | MaskOCR (ViT-B) | Accuracy (Scene)73.9 | 9 | 3mo ago | |
| OmniDoc | MinerU2.5-Pro | Edit Distance0.064 | 8 | 15d ago | |
| DocLayNet | Edit Distance8 | 8 | 15d ago | ||
| Turb-text (test) | DATUM | CRNN Accuracy93.55 | 7 | 3mo ago | |
| IIIT5K | CA-FCN | Accuracy (small)99.8 | 6 | 3mo ago | |
| ICDAR robust reading 2013 (test) | DictNet | Accuracy90.8 | 6 | 3mo ago | |
| RoadText-1K | Transcription Rate33.3 | 5 | 23h ago |
