Synthetic Document Generator for Annotation-free Layout Recognition

TL;DR

This paper introduces a synthetic document generator that creates realistic, labeled document images for training layout recognition models, reducing the need for costly manual annotations.

Contribution

It presents a novel hierarchical generative process using Bayesian Networks and stochastic templates to produce diverse, realistic synthetic documents for layout detection training.

Findings

Synthetic documents enable training models that perform comparably to those trained on real data.

The generative approach captures complex layout variations effectively.

The method reduces dependency on expensive manual annotations.

Abstract

Analyzing the layout of a document to identify headers, sections, tables, figures etc. is critical to understanding its content. Deep learning based approaches for detecting the layout structure of document images have been promising. However, these methods require a large number of annotated examples during training, which are both expensive and time consuming to obtain. We describe here a synthetic document generator that automatically produces realistic documents with labels for spatial positions, extents and categories of the layout elements. The proposed generative process treats every physical component of a document as a random variable and models their intrinsic dependencies using a Bayesian Network graph. Our hierarchical formulation using stochastic templates allow parameter sharing between documents for retaining broad themes and yet the distributional characteristics produces visually unique samples, thereby capturing complex and diverse layouts. We empirically illustrate that a deep layout detection model trained purely on the synthetic documents can match the performance of a model that uses real documents.