Thai music emotion recognition based on western music

Abstract

Music emotion recognition is the music emotion detected from people's annotations. In this paper, the Thai music was the evaluated set of a system based on western music training settings. By using valence-arousal values, multiple linear regression, k-nearest neighbours to represent the emotional annotations from the music. We used valence and energy(arousal) from Spotify API to the investigated emotion of Thai music. As a result, the Thai music emotion according to the western music criteria could be understood. The highest f-measure of Thai music from multiple linear regression All feature was 41% and the f-measure of western music from multiple linear regression without tempo feature was 51 %, which are very different because All feature in western music is low efficiency than other models.