@InProceedings{PauTorras2022,
author="Pau Torras
and Arnau Baro
and Alicia Fornes
and Lei Kang",
title="Improving Handwritten Music Recognition through Language Model Integration",
booktitle="4th International Workshop on Reading Music Systems (WoRMS2022)",
year="2022",
pages="42--46",
optkeywords="optical music recognition",
optkeywords="historical sources",
optkeywords="diversity",
optkeywords="music theory",
optkeywords="digital humanities",
abstract="Handwritten Music Recognition, especially in the historical domain, is an inherently challenging endeavour; paper degradation artefacts and the ambiguous nature of handwriting make recognising such scores an error-prone process, even for the current state-of-the-art Sequence to Sequence models. In this work we propose a way of reducing the production of statistically implausible output sequences by fusing a Language Model into a recognition Sequence to Sequence model. The idea is leveraging visually-conditioned and context-conditioned output distributions in order to automatically find and correct any mistakes that would otherwise break context significantly. We have found this approach to improve recognition results to 25.15 SER (\%) from a previous best of 31.79 SER (\%) in the literature.",
optnote="DAG; 600.121; 600.162; 602.230",
optnote="exported from refbase (http://refbase.cvc.uab.es/show.php?record=3735), last updated on Thu, 27 Apr 2023 14:55:49 +0200",
file=":http://refbase.cvc.uab.es/files/TBF2022.pdf:PDF"
}