PubLayNet: Largest Dataset Ever for Document Layout Analysis

Zhong, Xu; Tang, Jianbin; Yepes, Antonio Jimeno

doi:10.1109/icdar.2019.00166

Cited by 369 publications

(277 citation statements)

References 13 publications

Supporting

Mentioning

264

Contrasting

Unclassified

Order By: Relevance

“…More on this in sections "Investment documents datasets" and "Title hierarchization" 7. an exhaustive study reporting on the usage of prospectuses confirms this: https ://morec arrot .com/wp-conte nt/uploa ds/2019/10/MC_ Prosp ectus _Study Repor tFina l_23oct 19.pdf with MS Word used in 92% of the cases 8. for prospectuses, see https ://www.amf-franc e.org/en_US/Formu laire s-et-decla ratio ns/OPCVM -et-fonds -d-inves tisse ment/OPCVM / Plan-type-du-prosp ectus 0.…”

mentioning

confidence: 78%

“…More recently, [8] released the largest dataset to benchmark layout structure extraction algorithms, called Pub-LayNet. With over 1 million scientific articles from the ArXiv.org website, the authors train a neural network on a page segmentation task to detect text, titles, lists, tables and figures from images.…”

Section: Text and Layout Extraction From Imagesmentioning

confidence: 99%

“…French documents never include an embedded TOC page whereas English documents do, but this parsable TOC page is usually highly incomplete as only the higher-level titles are displayed. Due to the presence of a template for the French investment documents 8 , we noticed key differences between French and English documents. In French documents, the titles usually contain the same words but the phrasing can differ from a document to another.…”

Section: Investment Documents Datasetsmentioning

confidence: 99%

See 2 more Smart Citations

Automatic Table-of-Contents Generation for Efficient Information Access

et al. 2020

View full text Add to dashboard Cite

Purpose This paper presents a novel neural-based approach, applicable to any searchable PDF document that first detects the titles and then hierarchically orders them using a sequence labelling approach to generate automatically the Table of Contents (TOC). A TOC signals the main divisions and subdivisions of a document to assist with navigation and information localisation. Methods Unlike previous methods, we do not assume the presence of parsable TOC pages in the document but infer the TOC from a data-driven analysis of sections titles, their order and their depth. Results We offer an exhaustive analysis of the proposed model and evaluate it on French and English using documents from the financial domain, which we release to increase community's interest. We compare this model to state-of-the-art approaches and show its superiority in multiple experiments. Conclusions The approach described in this paper can easily be adapted to other domains and documents and its application to the analysis of financial prospectuses will be strengthened by the release of datasets. The TOC generation algorithms used in this paper obtain state-of-the-art results and provide strong baselines for future work.

show abstract

mentioning

confidence: 78%

Section: Text and Layout Extraction From Imagesmentioning

confidence: 99%

Section: Investment Documents Datasetsmentioning

confidence: 99%

See 1 more Smart Citation

Automatic Table-of-Contents Generation for Efficient Information Access

et al. 2020

View full text Add to dashboard Cite

show abstract

“…Page segmentation. Also called document layout analysis, page segmentation is an active research area with numerous competitions [1]- [4] and datasets [5]- [8]. They usually consider many semantic categories (e.g., caption, paragraph, title) and split text regions at paragraph level.…”

Section: Related Workmentioning

confidence: 99%

“…Training deep networks for both page and text line segmentation requires large amounts of data. For modern documents, Yang et al [7] and Zhong et al [8] proposed synthetic document generation engines based on modern formats (respectively Latex and PDF) yielding to large-scale and heterogeneous document datasets. However, these documents are too simple to train a model that perform well on historical documents.…”

Section: Related Workmentioning

confidence: 99%

docExtractor: An off-the-shelf historical document element extraction

Monnier

Aubry

2020

2020 17th International Conference on Frontiers in Handwriting Recognition (ICFHR)

View full text Add to dashboard Cite

We present docExtractor, a generic approach for extracting visual elements such as text lines or illustrations from historical documents without requiring any real data annotation. We demonstrate it provides high-quality performances as an offthe-shelf system across a wide variety of datasets and leads to results on par with state-of-the-art when fine-tuned. We argue that the performance obtained without fine-tuning on a specific dataset is critical for applications, in particular in digital humanities, and that the line-level page segmentation we address is the most relevant for a general purpose element extraction engine. We rely on a fast generator of rich synthetic documents and design a fully convolutional network, which we show to generalize better than a detection-based approach. Furthermore, we introduce a new public dataset dubbed IlluHisDoc dedicated to the fine evaluation of illustration segmentation in historical documents.

show abstract

Enhancing optical character recognition: Efficient techniques for document layout analysis and text line detection

Fateh,

Abolghasemi

2023

Engineering Reports

View full text Add to dashboard Cite

In recent years, automatic document and text analysis has gained significant importance, driven by advancements in optical character recognition (OCR) technology and the need for efficient processing of large volumes of printed or handwritten documents. This article specifically focuses on document layout analysis (DLA) and text line detection (TLD), both of which are crucial components of OCR systems. Our objective is to develop an effective method for extracting both textual and non‐textual regions, addressing challenges unique to the Persian (and Persian‐like) language(s). In the DLA stage, we employ deep learning models and a voting system to accurately determine the regions of interest. Additionally, we introduce methods such as optimum font size concepts, angle correction, and a line curvature elimination algorithm in the TLD process to enhance OCR accuracy. Comparative evaluations against state‐of‐the‐art methods demonstrate the superiority of our approach, showcasing a 2.8% improvement in the accuracy of Tesseract‐OCR 5.1.0 (a well‐established commercial OCR system) on the official Iranian newspapers dataset. These findings underscore the importance of addressing DLA and TLD challenges to advance OCR technology for Persian language documents and provide a solid foundation for future research in this domain.

show abstract

PubLayNet: Largest Dataset Ever for Document Layout Analysis

Cited by 369 publications

References 13 publications

Automatic Table-of-Contents Generation for Efficient Information Access

Automatic Table-of-Contents Generation for Efficient Information Access

docExtractor: An off-the-shelf historical document element extraction

Enhancing optical character recognition: Efficient techniques for document layout analysis and text line detection

Contact Info

Product

Resources

About