| --- |
| license: other |
| license_name: per-source-licences |
| license_link: https://github.com/OpenCR-lab/BanglaDocBench#licence |
| language: |
| - bn |
| pretty_name: BanglaDocBench |
| task_categories: |
| - image-to-text |
| tags: |
| - bangla |
| - bengali |
| - documents |
| - ocr |
| - document-parsing |
| - benchmark |
| --- |
| |
| # BanglaDocBench |
|
|
| **A document-parsing benchmark for Bangla: 500 to 1,000 pages chosen for diversity from 19.9 |
| million.** |
|
|
| > **Status: in development. No benchmark data has been released yet.** This repository reserves |
| > the name and will hold the benchmark when it is finished. |
|
|
| BanglaDocBench will measure how well OCR and document-parsing systems read Bangla documents: |
| books, newspapers, government notices, textbooks, forms, old prints and modern typesetting. It |
| follows the approach of benchmarks such as |
| [OmniDocBench](https://huggingface.co/datasets/opendatalab/OmniDocBench), for Bangla script. |
|
|
| ## How the pages are chosen |
|
|
| The pages are drawn from **[BanglaDocIndex](https://huggingface.co/datasets/OpenCR-lab/BanglaDocIndex)**, |
| a public index of 19,873,418 Bangla document pages with an EmbeddingGemma 2 embedding for every |
| page. |
|
|
| 1. **Clustering.** The page embeddings are clustered to map the kinds of pages in the pool. |
| 2. **Selection.** Pages are sampled across clusters for diversity, with a fixed random seed. |
| 3. **Annotation.** Each selected page is annotated (text, reading order and layout). |
|
|
| The cluster labels, the selected page IDs, the seed and the selection code will be published with |
| the benchmark, and the selected pages will be marked in BanglaDocIndex so that training sets can |
| exclude them. |
|
|
| ## Licence |
|
|
| - **Code:** Apache-2.0, at [github.com/OpenCR-lab/BanglaDocBench](https://github.com/OpenCR-lab/BanglaDocBench). |
| - **Annotations:** CC-BY-4.0 (planned). |
| - **Page images are not relicensed.** Each page image remains under the licence or terms of its |
| source document. The source, the source URL and the source's own rights statement are recorded |
| for every page. Many sources state no licence; copyright in those pages stays with their rights |
| holders, and they are included for research and evaluation only. Anyone using the images is |
| responsible for following the terms of each source. |
| - **Removal requests:** if you hold the rights to a page and want it removed, open an issue at |
| [github.com/OpenCR-lab/BanglaDocBench/issues](https://github.com/OpenCR-lab/BanglaDocBench/issues) |
| with the page ID or source URL. |
|
|
