mlsanityData to 14 Oct 2026

Page Layout Analysis of Text-heavy Historical Documents: a Comparison of Textual and Visual Approaches

arXiv 2212.13924 ↗First posted 12 Dec 2022cs.IR cs.AI cs.CLcited by 7

Not on the map. Why?

Abstract (arXiv)

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.

Key facts. Cited by 7 papers in the map (data through 14 Oct 2026). Builds on 8 of its 20 references that are in the map. First posted on arXiv 12 Dec 2022.

Builds on

8 of 20 references.

References · 8

  1. 2018BERT: Pre-training of Deep Bidirectional Transformers for Language Understandingcited by 40,399
  2. 2019RoBERTa: A Robustly Optimized BERT Pretraining Approachcited by 13,372
  3. 2022LayoutLMv3: Pre-training for Document AI with Unified Text and Image Maskingcited by 485
  4. 2019The VIA Annotation Software for Images, Audio and Videocited by 167
  5. 2022DiT: Self-supervised Pre-training for Document Image Transformercited by 152
  6. 2017Learning to Extract Semantic Structure from Documents Using Multimodal Fully Convolutional Neural Networkcited by 92
  7. 2020Combining Visual and Textual Features for Semantic Segmentation of Historical Newspaperscited by 10
  8. 2022You Actually Look Twice At it (YALTAi): using an object detection approach instead of region segmentation within the Kraken enginecited by 2

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2023: 1; 2024: 1; 2025: 2; 2026: 3.

7 papers.

  1. 2023Yes but.. Can ChatGPT Identify Entities in Historical Documents?cited by 10
  2. 2026Structure-Aware Text Recognition for Ancient Greek Critical Editionscited by 5
  3. 2025Reading the unreadable: Creating a dataset of 19th century English newspapers using image-to-text language modelscited by 3
  4. 2024Diachronic Document Dataset for Semantic Layout Analysiscited by 1
  5. 2025Studying Illustrations in Manuscripts: An Efficient Deep-Learning Approachcited by 1
  6. 2026Semantic-Guided Reading Order Reconstruction in Historical Armenian Newspapers with LLMscited by 1
  7. 2026The Patrologia Graeca Corpus: OCR, Annotation, and Open Release of Noisy Nineteenth-Century Polytonic Greek Editionscited by 0