Enhancing Object Detection in Ancient Documents with Synthetic Data Generation and Transformer-Based Models

TL;DR

This paper proposes a novel approach combining synthetic data generation and transformer-based models to improve object detection accuracy in ancient documents, addressing challenges posed by low quality and complex details.

Contribution

It introduces a synthetic dataset creation method and integrates visual feature maps into object detection models for better symbol and element recognition.

Findings

Significant reduction in false positives.

Enhanced detection precision in ancient documents.

Improved analysis capabilities in Paleography.

Abstract

The study of ancient documents provides a glimpse into our past. However, the low image quality and intricate details commonly found in these documents present significant challenges for accurate object detection. The objective of this research is to enhance object detection in ancient documents by reducing false positives and improving precision. To achieve this, we propose a method that involves the creation of synthetic datasets through computational mediation, along with the integration of visual feature extraction into the object detection process. Our approach includes associating objects with their component parts and introducing a visual feature map to enable the model to discern between different symbols and document elements. Through our experiments, we demonstrate that improved object detection has a profound impact on the field of Paleography, enabling in-depth analysis and fostering a greater understanding of these valuable historical artifacts.