Metrical-accent Aware Vocal Onset Detection in Polyphonic Audio

TitleMetrical-accent Aware Vocal Onset Detection in Polyphonic Audio
Publication TypeConference Paper
Year of PublicationIn Press
Conference NameThe 18th International Society for Music Information Retrieval Conference
AuthorsDzhambazov, G., Holzapfel A., Srinivasamurthy A., & Serra X.
Conference Start Date23/10/2017
Conference LocationSuzhou, China
AbstractThe goal of this study is the automatic detection of onsets of the singing voice in polyphonic audio recordings. Starting with a hypothesis that the knowledge of the current position in a metrical cycle (i.e. metrical accent) can improve the accuracy of vocal note onset detection, we propose a novel probabilistic model to jointly track beats and vocal note onsets. The proposed model extends a state of the art model for beat and meter tracking, in which a-priori probability of a note at a specific metrical accent interacts with the probability of observing a vocal note onset. We carry out an evaluation on a varied collection of multi-instrument datasets from two music traditions (English popular music and Turkish makam) with different types of metrical cycles and singing styles. Results confirm that the proposed model reasonably improves vocal note onset detection accuracy compared to a baseline model that does not take metrical position into account.
preprint/postprint documenthttps://arxiv.org/pdf/1707.06163.pdf
Additional material: 

Code repositories: https://github.com/georgid/pypYIN for section 5.2

and https://github.com/georgid/madmom for section 5.3

Datasets companion page: http://compmusic.upf.edu/node/345

 

 

intranet