Florian Henkel, Rainer Kelz, Gerhard Widmer,
"Learning to Read and Follow Music Incomplete Score Sheet Images"
: In Proceedings of the 21st International Society for MusicInformation Retrieval Conference, 2020, 7-2020
Learning to Read and Follow Music Incomplete Score Sheet Images
Sprache des Titels:
In Proceedings of the 21st International Society for MusicInformation Retrieval Conference, 2020
This paper addresses the task of score following in sheetmusic given as unprocessed images. While existing workeither relies on OMR software to obtain a computer-readable score representation, or crucially relies on pre-pared sheet image excerpts, we propose the first systemthat directly performs score following in full-page, com-pletely unprocessed sheet images. Based on incoming au-dio and a given image of the score, our system directly pre-dicts the most likely position within the page that matchesthe audio, outperforming current state-of-the-art image-based score followers in terms of alignment precision. Wealso compare our method to an OMR-based approach andempirically show that it can be a viable alternative to sucha system