Walshaw, Chris ORCID: 0000-0003-0253-7779 (2015) Multilevel melodic matching. In: Proceedings of the 5th International Workshop on Folk Music Analysis. Association Dirac, Paris, pp. 130-137. ISBN 9791095209003

Preview

PDF (Open Access Conference Article) 14023_WALSHAW_Multilevel_Melodic_Matching_2015.pdf - Published Version Download (5MB)

Abstract

This paper describes a multilevel algorithm for matching tunes when performing inexact searches in symbolic mu-sical data. The basis of the algorithm is straightforward: initially each tune in the search database is normalised and quantised and then recursively coarsened, typically by removing weaker off-beats, until the tune is reduced to a skeleton representation with just one note per bar. The same process is applied to the search query and melodic matching between query and data can then take place at every level. The algorithm implemented here uses the longest common substring algorithm at each level, but in principle a variety of similarity measures could be used. The multilevel framework allows inexact matches to occur by identifying similarities at course levels and is also exploited with the use of early termination heuristics at coarser levels, both to reduce computational complexity and to enhance the matching qualitatively. Experimenta-tion demonstrates the effectiveness of the approach for inexact melodic searches within a corpus of tunes.

Item Type:	Conference Proceedings
Title of Proceedings:	Proceedings of the 5th International Workshop on Folk Music Analysis
Additional Information:	The 5th International Workshop on Folk Music Analysis (http://fma2015.sciencesconf.org) was organized by the LAM Team (Lutheries – Acoustique – Musique) from the Institute D’Alembert CNRS UMR 7190.   This 2015 session was hosted by the University Pierre and Marie Curie (www.upmc.fr), Paris, France, on 10 - 12 June 2015.
Uncontrolled Keywords:	cultural informatics; music similarity;
Subjects:	M Music and Books on Music > MT Musical instruction and study
Faculty / School / Research Centre / Research Group:	Faculty of Engineering & Science > School of Computing & Mathematical Sciences (CMS) Faculty of Engineering & Science
Related URLs:	Organisation
Last Modified:	04 Mar 2022 13:07
URI:	http://gala.gre.ac.uk/id/eprint/14023

Actions (login required)

View Item

Downloads