mlsanityData to 7 Oct 2026

When Do VLMs Help Arabic Manuscript OCR? A Cross-Dataset Study

arXiv 2608.22366 ↗First posted 23 Aug 2026cs.CVcited by 0

On the map: Multimodal LLMs / vision-language models → Document understanding and multimodal RAG

Abstract (arXiv)

Vision-language models (VLMs) are increasingly being used for document understanding, yet their role in Arabic and Islamic manuscript recognition remains underexplored. To address such a gap in this paper, we evaluate traditional OCR, general-purpose VLMs, Arabic-specialized VLMs, and OCR-conditioned VLM correction across eight Arabic text datasets spanning historical manuscripts, aged printed books, clean print, multi-domain documents, and handwriting. The results show that no single approach dominates across setups. On line-level historical manuscripts, VLMs are close to Tesseract; on page-level manuscript images, they perform better; and in several settings, an OCR-conditioned corrector improves over both standalone OCR and standalone VLMs. The central finding is an OCR-prior recoverability principle: OCR conditioning helps when the OCR output remains visually and textually recoverable, providing anchors that the VLM can refine against the image. It improves recognition on aged print, clean print, mixed-domain Arabic, and some Naskh manuscripts, but degrades performance when the prior is script-mismatched or systematically misleading, as in Maghribi manuscripts and realistic student handwriting. Additional diagnostics show that Arabic VLM-OCR is sensitive to diacritics, preprocessing, generation budget, and repetition loops. These findings support an adaptive OCR-VLM workflow that routes pages according to script, OCR-prior recoverability, length diagnostics, and failure-mode indicators.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Placed in Document understanding and multimodal RAG, one of 16 directions in Multimodal LLMs / vision-language models. Builds on 22 of its 38 references that are in the map. First posted on arXiv 23 Aug 2026.

Builds on

22 of 38 references.

Closest ancestors · 2

Same direction, newest first

  1. 2025Baseer: A Vision-Language Model for Arabic Document-to-Markdown OCR17 Sep
  2. 2025Multimodal LLMs for OCR, OCR Post-Correction, and Named Entity Recognition in Historical Documents1 Apr

Other references · 20

  1. 2021OCR-free Document Understanding Transformercited by 400
  2. 2024General OCR Theory: Towards OCR-2.0 via a Unified End-to-end Modelcited by 174
  3. 2025PalmX 2025: The First Shared Task on Benchmarking LLMs on Arabic and Islamic Culturecited by 20
  4. 2025KITAB-Bench: A Comprehensive Multi-Domain Benchmark for Arabic OCR and Document Understandingcited by 14
  5. 2024Muharaf: Manuscripts of Handwritten Arabic Dataset for Cursive Text Recognitioncited by 13
  6. 2019BADAM: A Public Dataset for Baseline Detection in Arabic-script Manuscriptscited by 12
  7. 2025Benchmarking Large Language Models for Handwritten Text Recognitioncited by 11
  8. 2025Early evidence of how LLMs outperform traditional systems on OCR/HTR tasks for historical recordscited by 10
  9. 2025OCR Error Post-Correction with LLMs in Historical Documents: No Free Lunchescited by 10
  10. 2024Native vs Non-Native Language Prompting: A Comparative Analysiscited by 9
  11. 2024HATFormer: Historic Handwritten Arabic Text Recognition with Transformerscited by 9
  12. 2025CHURRO: Making History Readable with an Open-Weight Large Vision-Language Model for High-Accuracy, Low-Cost Historical Text Recognitioncited by 9
8 more
  1. 2026From RAG to Agentic RAG for Faithful Islamic Question Answeringcited by 9
  2. 2023Advancements and Challenges in Arabic Optical Character Recognition: A Comprehensive Surveycited by 8
  3. 2026Fanar 2.0: Arabic Generative AI Stackcited by 8
  4. 2024Qalam : A Multimodal LLM for Arabic Optical Character and Handwriting Recognitioncited by 7
  5. 2024Reference-Based Post-OCR Processing with LLM for Precise Diacritic Text in Historical Document Recognitioncited by 5
  6. 2025SARD: A Large-Scale Synthetic Arabic OCR Dataset for Book-Style Text Recognitioncited by 5
  7. 2025QARI-OCR: High-Fidelity Arabic Text Recognition through Multimodal Large Language Model Adaptationcited by 5
  8. 2025Evaluating LLMs for Historical Document OCR: A Methodological Framework for Digital Humanitiescited by 5

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2026RefLAM: A Reference-Grounded Line Annotation Pipeline for Historical Arabic Manuscripts
  2. 2024DocVLM: Make Your VLM an Efficient Reader
  3. 2026Do VLMs Read or Rewrite? On Transcription Faithfulness in Vision-Language Models
  4. 2026AtlasOCR: Building the First Open-Source Darija OCR Model with Vision Language Models
  5. 2026Can OCR-VLMs Read Devanagari? A Stress-Test Benchmark and Post-Correction Study
  6. 2026A Survey of OCR Evaluation Methods and Metrics and the Invisibility of Historical Documents
  7. 2026When Simpler Is Better: Evaluating Translation Pipelines for Medieval Latin Manuscripts
  8. 2026Reading or Guessing? Visual Grounding Failures of Vision-Language Models for OCR in Ancient Greek Editions