We present a new dataset for form understanding in noisy scanned documents (FUNSD) that aims at extracting and structuring the textual content of forms. The dataset comprises 199 real, fully annotated, scanned forms. The documents are noisy and vary widely in appearance, making form understanding (FoUn) a challenging task. The proposed dataset can be used for various tasks, including text detection, optical character recognition, spatial layout analysis, and entity labeling/linking. To the best of our knowledge, this is the first publicly available dataset with comprehensive annotations to address FoUn task.
Title
FUNSD: A Dataset for Form Understanding in Noisy Scanned Documents
Published in
2019 International Conference On Document Analysis And Recognition Workshops (Icdarw) And 2Nd International Workshop On Open Services And Tools For Document Analysis (Ost), Vol 2
Series
Proceedings of the International Conference on Document Analysis and Recognition
Pages
1-6
Conference
15th IAPR International Conference on Document Analysis and Recognition (ICDAR) / 2nd International Workshop on Open Services and Tools for Document Analysis (OST), Sep 21-25, 2019, Sydney, AUSTRALIA
Date
2019-01-01
Publisher
New York, IEEE
ISSN
1520-5363
ISBN
978-1-7281-5054-3
Record creation date
2020-03-26