Generating Synthetic Handwritten Historical Documents With OCR Constrained GANs

TL;DR

This paper introduces a novel framework using OCR-constrained GANs to generate high-quality synthetic historical documents with accurate ground truth, aiding the development of supervised Document Image Analysis methods.

Contribution

The authors propose a two-step style transfer approach that creates realistic synthetic historical documents from templates, requiring no expert knowledge and preserving layout and text.

Findings

Synthetic documents improve pre-training performance over baselines.

The method produces high-quality, realistic historical document images.

Large labeled datasets can be generated for training without manual annotation.

Abstract

We present a framework to generate synthetic historical documents with precise ground truth using nothing more than a collection of unlabeled historical images. Obtaining large labeled datasets is often the limiting factor to effectively use supervised deep learning methods for Document Image Analysis (DIA). Prior approaches towards synthetic data generation either require expertise or result in poor accuracy in the synthetic documents. To achieve high precision transformations without requiring expertise, we tackle the problem in two steps. First, we create template documents with user-specified content and structure. Second, we transfer the style of a collection of unlabeled historical images to these template documents while preserving their text and layout. We evaluate the use of our synthetic historical documents in a pre-training setting and find that we outperform the baselines (randomly initialized and pre-trained). Additionally, with visual examples, we demonstrate a high-quality synthesis that makes it possible to generate large labeled historical document datasets with precise ground truth.