| --- |
| license: cc-by-4.0 |
| library_name: pytorch |
| pipeline_tag: image-to-text |
| tags: |
| - handwritten-text-recognition |
| - document-layout-analysis |
| - htr |
| - ocr |
| - historical-documents |
| - read-2016 |
| --- |
| |
| # hand-read2016-page-compact |
|
|
| Part of **HAND: Unified Text–Layout Decoding for Handwritten Document Recognition** (Hamdan, Rahiche, Cheriet). |
| Code: [github.com/DocumentRecognitionModels/HAND-Decoding](https://github.com/DocumentRecognitionModels/HAND-Decoding) |
|
|
| The compact page model with one visual key/value projection shared by all eight decoder layers (6,112,612 parameters; 500,000 training samples). |
|
|
| - **Architecture:** fully convolutional encoder and eight-layer transformer decoder; one output |
| stream of characters and layout tokens (page, page number, section, annotation, body). |
| - **Parameters:** 6,112,612 plus 365,968 in the speculative draft heads (m = 5). |
| - **Input:** an RGB document image, resampled to 150 dpi. |
| - **Result reported in the paper:** READ 2016 single-page test (50 pages): CER 4.00 %, WER 15.74 %, LOER 0.0349, mAP-CER 0.9365. Trained for fewer samples than the full page model, so its error rates are not comparable with it. |
| - **Scope:** Reads one page. |
|
|
| Verified before release: decoding the stored example image with these weights reproduces the |
| prediction recorded in the repository token for token; with the draft heads, speculative decoding reproduces greedy decoding. |
|
|
| ```python |
| # git clone https://github.com/DocumentRecognitionModels/HAND-Decoding && cd HAND-Decoding && pip install -r requirements.txt |
| import sys; sys.path.insert(0, "release") |
| from hand_release.inference import HANDRecognizer |
| |
| model = HANDRecognizer.from_pretrained("MHamdan/hand-read2016-page-compact", kv_cache=True, speculative=True) |
| page = model.read("page.jpg", source_dpi=300) # 300 for a raw scan, 150 if already resampled |
| print(page.text) # transcription; page.raw keeps the layout tokens inline |
| ``` |
|
|
| ## Licence and attribution |
|
|
| These weights are released under the Creative Commons Attribution 4.0 International licence |
| (CC BY 4.0, https://creativecommons.org/licenses/by/4.0/). They are Adapted Material of: |
|
|
| - **Pretrained Document Attention Network for Handwritten Text Recognition**, Denis Coquenet, |
| Zenodo, DOI 10.5281/zenodo.7244382, CC BY 4.0, file `fcn_read_2016_line_syn.pt` |
| (sha256 557d6c349131b113c1e316effdb028eff9b37bf39a48d0689125a3783b86a7c3). Its encoder weights |
| initialised this model, which was then trained further on READ 2016 page images and synthetic |
| pages. The model was trained on READ 2016 (Sánchez et al., Zenodo 10.5281/zenodo.1297399, |
| CC BY 4.0). |
|
|
| The weights are provided as-is, without warranty. The code that loads them is MIT-licensed, with |
| DAN-derived files under CeCILL-C; see `release/NOTICE.md` in |
| https://github.com/DocumentRecognitionModels/HAND-Decoding. |
|
|
| ## Citation |
|
|
| The paper is under review. Until it is published, please cite the preprint, which appeared under an earlier title: |
|
|
| ```bibtex |
| @article{hamdan2024hand, |
| title = {{HAND}: Hierarchical Attention Network for Multi-Scale Handwritten |
| Document Recognition and Layout Analysis}, |
| author = {Hamdan, Mohammed and Rahiche, Abderrahmane and Cheriet, Mohamed}, |
| journal = {arXiv preprint arXiv:2412.18981}, |
| year = {2024}, |
| url = {https://arxiv.org/abs/2412.18981} |
| } |
| ``` |
|
|
