Page Layout Analysis of Text-heavy Historical Documents: a Comparison of Textual and Visual Approaches
Preprint
Abstract
Page layout analysis is a fundamental step in document processing which enables to segment a page into regions of interest. With highly complex layouts and mixed scripts, scholarly commentaries are text-heavy documents which remain challenging for state-of-the-art models. Their layout considerably varies across editions and their most important regions are mainly defined by semantic rather than graphical characteristics such as position or appearance. This setting calls for a comparison between textual, visual and hybrid approaches. We therefore assess the performances of two transformers (LayoutLMv3 and RoBERTa) and an objection-detection network (YOLOv5). If results show a clear advantage in favor of the latter, we also list several caveats to this finding. In addition to our experiments, we release a dataset of ca. 300 annotated pages sampled from 19th century commentaries.Keywords
All Related Versions
This publication has 14 references indexed in Scilit:
- LayoutLMv3: Pre-training for Document AI with Unified Text and Image Masking2022
- DiT: Self-supervised Pre-training for Document Image Transformer2022
- BiblIA - a General Model for Medieval Hebrew Manuscripts and an Open Annotated Dataset2021
- The VIA Annotation Software for Images, Audio and Video2019
- ICDAR2019 Competition on Historical Book Analysis - HBA20192019
- The Free First Thousand Years of Greek2019
- ICDAR2017 Competition on Layout Analysis for Challenging Medieval Manuscripts2017
- Learning to Extract Semantic Structure from Documents Using Multimodal Fully Convolutional Neural Networks2017
- A comprehensive survey of mostly textual document segmentation algorithms since 2008Pattern Recognition, 2017
- Page Segmentation Techniques in Document Analysis2014