A simple fusion method of state and sequence segmentation for music structure discovery

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

Methods for music structure segmentation are based on strong assumptions on the acoustical properties of structural segments. These assumptions relate to the novelty, homogeneity, repetition and/or regularity of the content. Each of these assumptions provide a different perspective on the music piece. These assumptions are however often considered separately in the methods. In this paper we propose a method for estimating the music structure segmentation based on the fusion of the novelty and repetition assumptions. This combination of different perspectives on the music pieces allows to generate more coherent acoustic segments and strongly improves the final music structure segmentation’s performance.

Original languageEnglish
Title of host publicationProceedings of the 14th International Society for Music Information Retrieval Conference, ISMIR 2013
EditorsAlceu de Souza Britto, Fabien Gouyon, Simon Dixon
PublisherInternational Society for Music Information Retrieval
Pages257-262
Number of pages6
ISBN (Electronic)9780615900650
Publication statusPublished - 1 Jan 2013
Externally publishedYes
Event14th International Society for Music Information Retrieval Conference, ISMIR 2013 - Curitiba, Brazil
Duration: 4 Nov 20138 Nov 2013

Publication series

NameProceedings of the 14th International Society for Music Information Retrieval Conference, ISMIR 2013

Conference

Conference14th International Society for Music Information Retrieval Conference, ISMIR 2013
Country/TerritoryBrazil
CityCuritiba
Period4/11/138/11/13

Fingerprint

Dive into the research topics of 'A simple fusion method of state and sequence segmentation for music structure discovery'. Together they form a unique fingerprint.

Cite this