Deciphering the Underserved: Benchmarking LLM OCR for Low-Resource Scripts
It addresses accessibility gaps in text digitization for underserved languages, but is incremental as it highlights existing limitations and calls for further work.
This study tackled the problem of using Large Language Models (LLMs) like GPT-4o for Optical Character Recognition (OCR) in low-resource scripts such as Urdu, Albanian, and Tajik, and found that zero-shot LLM-based OCR has significant limitations, especially for linguistically complex scripts, based on a dataset of 2,520 images with controlled variations.
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.