The extraction of popular music chorus via structural content analysis

Chia Hung Yeh*, Hung Hsuan Lin

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Automatic chorus extraction from popular music is an interesting topic. Chorus extraction facilitates a user to quickly and efficiently preview selections from a large music database and also an essential preprocessing step for further purposes such as indexing and search. In this paper, a framework for extracting chorus from popular music based on structural content analysis is proposed. The repetitive structure is characterized via music color representation called colormap. A colormap of a song based on mapping different frequency bands to color space is used to find repeating patterns. This representation efficiently reveals the relationship of music structure. MFCCs (Mel Frequency Cepstral Coefficients) are employed to identify chorus and verse of a song according the structure information and music domain knowledge. Experimental results show that the performance of the proposed method over a sizable popular song database.

Original languageEnglish
Title of host publicationProceedings of the 33rd Annual Conference of the IEEE Industrial Electronics Society, IECON
Pages2532-2536
Number of pages5
DOIs
Publication statusPublished - 2007
Externally publishedYes
Event33rd Annual Conference of the IEEE Industrial Electronics Society, IECON - Taipei, Taiwan
Duration: 2007 Nov 52007 Nov 8

Publication series

NameIECON Proceedings (Industrial Electronics Conference)

Other

Other33rd Annual Conference of the IEEE Industrial Electronics Society, IECON
Country/TerritoryTaiwan
CityTaipei
Period2007/11/052007/11/08

Keywords

  • Chorus detection
  • Colormap
  • MFCC
  • Music structure analysis

ASJC Scopus subject areas

  • Control and Systems Engineering
  • Electrical and Electronic Engineering

Fingerprint

Dive into the research topics of 'The extraction of popular music chorus via structural content analysis'. Together they form a unique fingerprint.

Cite this