Exploiting Context in Handwriting Recognition Using Trainable Relaxation Labeling

Sara Ferro, Sara Ferro, Alessandro Torcinovich, Arianna Traviglia, Marcello Pelillo, Marcello Pelillo

2023

Abstract

Handwriting Text Recognition (HTR) is a fast-moving research topic in computer vision and machine learning domains. Many models have been introduced over the years, one of the most well-established ones being the Convolutional Recurrent Neural Network (CRNN), which combines convolutional feature extraction with recurrent processing of the visual embeddings. Such a model, however, presents some limitations such as a limited capability to account for contextual information. To counter this problem, we propose a new learning module built on top of the convolutional part of a classical CRNN model, derived from the relaxation labeling processes, which is able to exploit the global context reducing the local ambiguities and increasing the global consistency of the prediction. Experiments performed on three well-known handwritten recognition datasets demonstrate that the relaxation labeling procedures improve the overall transcription accuracy at both character and word levels.

Paper Citation

in Harvard Style

Ferro S., Torcinovich A., Traviglia A. and Pelillo M. (2023). Exploiting Context in Handwriting Recognition Using Trainable Relaxation Labeling. In Proceedings of the 12th International Conference on Pattern Recognition Applications and Methods - Volume 1: ICPRAM, ISBN 978-989-758-626-2, pages 574-581. DOI: 10.5220/0011849300003411

in Bibtex Style

@conference{icpram23,
author={Sara Ferro and Alessandro Torcinovich and Arianna Traviglia and Marcello Pelillo},
title={Exploiting Context in Handwriting Recognition Using Trainable Relaxation Labeling},
booktitle={Proceedings of the 12th International Conference on Pattern Recognition Applications and Methods - Volume 1: ICPRAM,},
year={2023},
pages={574-581},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0011849300003411},
isbn={978-989-758-626-2},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 12th International Conference on Pattern Recognition Applications and Methods - Volume 1: ICPRAM,
TI - Exploiting Context in Handwriting Recognition Using Trainable Relaxation Labeling
SN - 978-989-758-626-2
AU - Ferro S.
AU - Torcinovich A.
AU - Traviglia A.
AU - Pelillo M.
PY - 2023
SP - 574
EP - 581
DO - 10.5220/0011849300003411