FormulaNet: A Benchmark Dataset for Mathematical Formula Detection

One unsolved sub-task of document analysis is mathematical formula detection (MFD). Research by ourselves and others has shown that existing MFD datasets with inline and display formula labels are small and have insufficient labeling quality. There is therefore an urgent need for datasets with bette...

Full description

Saved in:
Bibliographic Details
Published in:IEEE access 2022, Vol.10, p.91588-91596
Main Authors: Schmitt-Koopmann, Felix M., Huang, Elaine M., Hutter, Hans-Peter, Stadelmann, Thilo, Darvishy, Alireza
Format: Article
Language:English
Subjects:
Annotations
Automatic annotation
Automation
Benchmark testing
dataset
Datasets
Deep learning
document analysis
Formulas (mathematics)
Labeling
Labelling
Labels
Mathematical analysis
mathematical formula detection
Mathematical models
Object recognition
page object detection
Pipelines
Text analysis
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:One unsolved sub-task of document analysis is mathematical formula detection (MFD). Research by ourselves and others has shown that existing MFD datasets with inline and display formula labels are small and have insufficient labeling quality. There is therefore an urgent need for datasets with better quality labeling for future research in the MFD field, as they have a high impact on the performance of the models trained on them. We present an advanced labeling pipeline and a new dataset called FormulaNet in this paper. At over 45k pages, we believe that FormulaNet is the largest MFD dataset with inline formula labels. Our experiments demonstrate substantially improved labeling quality for inline and display formulae detection over existing datasets. Additionally, we provide a math formula detection baseline for FormulaNet with an mAP of 0.754. Our dataset is intended to help address the MFD task and may enable the development of new applications, such as making mathematical formulae accessible in PDFs for visually impaired screen reader users.
ISSN:2169-3536
2169-3536
DOI:10.1109/ACCESS.2022.3202639