EUVIMEDEUVIMED entdecken
Crossref

DAN: A Segmentation-Free Document Attention Network for Handwritten Document Recognition

Denis Coquenet, Clement Chatelain, Thierry Paquet

IEEE Transactions on Pattern Analysis and Machine Intelligence · 01.07.2023 · 45 (7) · 8227-8243

DOI: 10.1109/tpami.2023.3235826 ↗

Kurz gefasst

Fachartikel in der Zeitschrift IEEE Transactions on Pattern Analysis and Machine Intelligence, erschienen am 01.07.2023. Verfasst von Denis Coquenet und 2 weiteren Autor:innen. Die Arbeit wurde bisher 121-mal zitiert (Crossref) und nennt 52 Literaturangaben. Gefördert von: French Defense Innovation Agency, Normandy.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Unconstrained handwritten text recognition is a challenging computer vision task. It is traditionally handled by a two-step approach, combining line segmentation followed by text line recognition. For the first time, we propose an end-to-end segmentation-free architecture for the task of handwritten document recognition: the Document Attention Network. In addition to text recognition, the model is trained to label text parts using begin and end tags in an XML-like fashion. This model is made up of an FCN encoder for feature extraction and a stack of transformer decoder layers for a recurrent token-by-token prediction process. It takes whole text documents as input and sequentially outputs characters, as well as logical layout tokens. Contrary to the existing segmentation-based approaches, the model is trained without using any segmentation label. We achieve competitive results on the READ 2016 dataset at page level, as well as double-page level with a CER of 3.43% and 3.70%, respectively. We also provide results for the RIMES 2009 dataset at page level, reaching 4.54% of CER. We provide all source code and pre-trained model weights at https://github.com/FactoDeepLearning/DAN.

Förderung

  • French Defense Innovation Agency
  • Normandy

Zitieren

APA

Coquenet, D., Chatelain, C., & Paquet, T. (2023). DAN: A Segmentation-Free Document Attention Network for Handwritten Document Recognition. IEEE Transactions on Pattern Analysis and Machine Intelligence, 45(7), 8227-8243. https://doi.org/10.1109/tpami.2023.3235826

Vancouver

Coquenet D, Chatelain C, Paquet T. DAN: A Segmentation-Free Document Attention Network for Handwritten Document Recognition. IEEE Transactions on Pattern Analysis and Machine Intelligence. 2023;45(7):8227-8243. doi:10.1109/tpami.2023.3235826

BibTeX
@article{Coquenet2023,
  author = {Coquenet, Denis and Chatelain, Clement and Paquet, Thierry},
  title = {DAN: A Segmentation-Free Document Attention Network for Handwritten Document Recognition},
  journal = {IEEE Transactions on Pattern Analysis and Machine Intelligence},
  year = {2023},
  volume = {45},
  number = {7},
  pages = {8227--8243},
  doi = {10.1109/tpami.2023.3235826}
}
Kostenlos starten →