# How Far Can Synthetic Data Take Thai OCR?

arXiv: https://arxiv.org/abs/2609.03595 · HTML: https://mlsanity.com/paper/2609.03595/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 14 of its 19 references that are in the map.
- First posted on arXiv 3 Sep 2026.

## Builds on (most cited first, 14 of 14)

- [Image-based table recognition: data, model, and evaluation](https://mlsanity.com/paper/1911.10683/) (2019)
- [OmniDocBench: Benchmarking Diverse PDF Document Parsing with Comprehensive Annotations](https://mlsanity.com/paper/2412.07626/) (2024)
- [DocLayNet: A Large Human-Annotated Dataset for Document-Layout Analysis](https://mlsanity.com/paper/2206.01062/) (2022)
- [CanvasVAE: Learning to Generate Vector Graphic Documents](https://mlsanity.com/paper/2108.01249/) (2021)
- [Optimized Table Tokenization for Table Structure Recognition](https://mlsanity.com/paper/2305.03393/) (2023)
- [PP-DocLayout: A Unified Document Layout Detection Model to Accelerate Large-Scale Data Construction](https://mlsanity.com/paper/2503.17213/) (2025)
- [PaddleOCR-VL-1.6: Expanding the Frontier of Document Parsing with Under-Optimized Region Refinement and Progressive Post-Training](https://mlsanity.com/paper/2606.03264/) (2026)
- [READoc: A Unified Benchmark for Realistic Document Structured Extraction](https://mlsanity.com/paper/2409.05137/) (2024)
- [PreP-OCR: A Complete Pipeline for Document Image Restoration and Enhanced OCR Accuracy](https://mlsanity.com/paper/2505.20429/) (2025)
- [Finetuning Vision-Language Models as OCR Systems for Low-Resource Languages: A Case Study of Manchu](https://mlsanity.com/paper/2507.06761/) (2025)
- [ThaiOCRBench: A Task-Diverse Benchmark for Vision-Language Understanding in Thai](https://mlsanity.com/paper/2511.04479/) (2025)
- [SEA-Vision: A Multilingual Benchmark for Comprehensive Document and Scene Text Understanding in Southeast Asia](https://mlsanity.com/paper/2603.15409/) (2026)
- [Cross-Lingual SynthDocs: A Large-Scale Synthetic Corpus for Any to Arabic OCR and Document Understanding](https://mlsanity.com/paper/2511.04699/) (2025)
- [Typhoon OCR: Open Vision-Language Model For Thai Document Extraction](https://mlsanity.com/paper/2601.14722/) (2026)

## Abstract (arXiv)

We investigate what makes synthetic OCR supervision transfer to real Thai documents and use the resulting insights to build Wayu-Paxa-OCR-Zero, a Thai OCR model adapted without OCR labels from real Thai document pages. Synthetic data provide exact labels at scale, but "realism" conflates source domain, page context, typography, spatial structure, and glyph variation. We disentangle these factors with a controlled document-reconstruction pipeline and evaluate each variant under page- and crop-level training on printed and handwritten Thai documents. Non-text context has little consistent effect, whereas typeface diversity, two-dimensional structure, and real handwriting glyphs improve transfer; moreover, source-domain matching depends on training granularity, with in-domain reconstruction approaching real printed supervision under page-level training (1.82% versus 1.31% median character error rate) but underperforming out-of-domain reconstruction under crop-level training (15.59% versus 5.52%). Guided by these findings, we adapt the 0.9B-parameter PaddleOCR-VL-1.6 into Wayu-Paxa-OCR-Zero using 45,723 synthetic pages: relative to its base checkpoint, it reduces median character error rate from 6.64% to 1.24% on printed pages and from 74.87% to 20.55% on handwriting and outperforms Typhoon OCR v1 7B on all five evaluation sets, showing that synthetic-only training can be competitive.