bbOCR: An Open-source Multi-domain OCR Pipeline for Bengali Documents

TL;DR

bbOCR is an open-source OCR system tailored for Bengali documents, utilizing novel models and synthetic datasets to improve digitization in low-resource language contexts.

Contribution

Introduces a scalable Bengali OCR system with new recognition models and synthetic datasets, enhancing document digitization for low-resource languages.

Findings

Outperforms existing Bengali OCR systems in evaluations.

Uses novel synthetic datasets for training and evaluation.

Provides open-source code and datasets for community use.

Abstract

Despite the existence of numerous Optical Character Recognition (OCR) tools, the lack of comprehensive open-source systems hampers the progress of document digitization in various low-resource languages, including Bengali. Low-resource languages, especially those with an alphasyllabary writing system, suffer from the lack of large-scale datasets for various document OCR components such as word-level OCR, document layout extraction, and distortion correction; which are available as individual modules in high-resource languages. In this paper, we introduce Bengali$.$AI-BRACU-OCR (bbOCR): an open-source scalable document OCR system that can reconstruct Bengali documents into a structured searchable digitized format that leverages a novel Bengali text recognition model and two novel synthetic datasets. We present extensive component-level and system-level evaluation: both use a novel diversified evaluation dataset and comprehensive evaluation metrics. Our extensive evaluation suggests that our proposed solution is preferable over the current state-of-the-art Bengali OCR systems. The source codes and datasets are available here: https://bengaliai.github.io/bbocr.