Deciphering the Underserved: Benchmarking LLM OCR for Low-Resource Scripts

Muhammad Abdullah Sohail, Salaar Masood, Hamza Iqbal
12/20/2024
cs.LGcs.CVeess.IV

No official code found. Best implementation: abdullahsohaill/CS6303-ResearchProject (0 stars, last commit 2025-12-30).

Abstract

This study investigates the potential of Large Language Models (LLMs), particularly GPT-4o, for Optical Character Recognition (OCR) in low-resource scripts such as Urdu, Albanian, and Tajik, with English serving as a benchmark. Using a meticulously curated dataset of 2,520 images incorporating controlled variations in text length, font size, background color, and blur, the research simulates diverse real-world challenges. Results emphasize the limitations of zero-shot LLM-based OCR, particularly for linguistically complex scripts, highlighting the need for annotated datasets and fine-tuned models. This work underscores the urgency of addressing accessibility gaps in text digitization, paving the way for inclusive and robust OCR solutions for underserved languages.

View on arXivView PDF

Code Implementations(1)

abdullahsohaill/CS6303-ResearchProjectCommunity100%

Benchmarking LLM OCR for Low-Resource Scripts

00Python, Jupyter NotebookOct 4, 20249 months ago

Similar Papers

Zero-shot OCR Accuracy of Low-Resourced Languages: A Comparative Analysis on Sinhala and Tamil

85% match2025

synthocr-gen: A synthetic ocr dataset generator for low-resource languages- breaking the data barrier

85% match2026

600k-ks-ocr: a large-scale synthetic dataset for optical character recognition in kashmiri script

84% match2026

VOLTAGE: A Versatile Contrastive Learning based OCR Methodology for ultra low-resource scripts through Auto Glyph Feature Extraction

84% match2025

OmniOCR: Generalist OCR for Ethnic Minority Languages

83% match1 repo32026

Low-Resource Language Processing: An OCR-Driven Summarization and Translation Pipeline

83% match2025

Cite this paper

@article{sohail2024deciphering,
  title  = {Deciphering the Underserved: Benchmarking LLM OCR for Low-Resource Scripts},
  author = {Muhammad Abdullah Sohail and Salaar Masood and Hamza Iqbal},
  year   = {2024},
  eprint = {2412.16119},
  archivePrefix = {arXiv},
  url    = {http://arxiv.org/abs/2412.16119v1}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/2412.16119.svg)](https://codeofpaper.com/paper/2412.16119)

Discussion