IAM
Unknown
13,353 handwritten text lines from 657 writers. Standard handwriting benchmark.
Benchmark Stats
SOTA History
wer
wer
Lower is better
| Rank | Model | Source | Score | Year | Paper |
|---|---|---|---|---|---|
| 1 | HTR-ConvText HTR-ConvText. Dec 2024. Table 2 IAM test set. | Community | 12.9 | 2024 | Source |
| 2 | HTR-VT HTR-VT. Li et al. 2024. Table 4 IAM test set. | Community | 14.9 | 2024 | Source |
| 3 | VAN Vertical Attention Network (VAN). WER from comparison tables in HTR-VT (2409.08573) and HTR-ConvText (2512.05021). IAM line-level. | Community | 16.3 | 2022 | Source |
cer
cer
Lower is better
| Rank | Model | Source | Score | Year | Paper |
|---|---|---|---|---|---|
| 1 | TrOCR-large TrOCR-large (BEiT-large + RoBERTa-large). Microsoft. Table 4 in Li et al. 2023. IAM line-level test split. SOTA at publication. | Community | 2.89 | 2023 | Source |
| 2 | TrOCR-base TrOCR-base (BEiT-base + RoBERTa-base). Microsoft. Table 4 in Li et al. 2023. IAM line-level test split. | Community | 3.42 | 2023 | Source |
| 3 | HTR-ConvText HTR-ConvText (ConvText block hybrid CNN+Transformer). DAIR-Group, Dec 2024. Table 2 IAM test set. SOTA among no-pretraining methods. | Community | 4 | 2024 | Source |
| 4 | HTR-VT HTR-VT (Vision Transformer for HTR, no pre-training or synthetic data). Li et al. 2024. Table 4 IAM test set. | Community | 4.7 | 2024 | Source |
| 5 | VAN Vertical Attention Network (VAN). Coquenet et al. IEEE TPAMI 2022. IAM line-level CER from comparison tables in HTR-VT (2409.08573) and HTR-ConvText (2512.05021). | Community | 5 | 2022 | Source |