Music emotion classification by audio signal analysis: analysis of self-selected music during game play

Don Knox, Gianna Cassidy, Scott Beveridge, Raymond A.R. Macdonald

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

82 Downloads (Pure)

Abstract

Music emotion classification algorithms seek to classify music files automatically by means of audio signal analysis. An overview of these methods is given, and an emotion classification algorithm is applied to the preferred music choices made by test subjects during a game play experiment. Results from the experiment are presented, in which test subjects were exposed to 3 sound conditions: preferred music, game soundtrack and experimenter-selected music. Obtained measures are heart rate, pedometer rate, game score, completion time and enjoyment. The preferred music choices from the experiment are analysed and classified according to mood cluster, valence and arousal. Obtained measures for these music classifications are discussed, as are the implications for automatic mood classification in choosing music for future experiments.

Original languageEnglish
Title of host publicationProceedings of the 10th International Conference on Music Perception and Cognition
Pages581-587
Number of pages7
Publication statusPublished - Aug 2008

Keywords

  • music
  • classification
  • analysis
  • emotion

Fingerprint Dive into the research topics of 'Music emotion classification by audio signal analysis: analysis of self-selected music during game play'. Together they form a unique fingerprint.

Cite this