Automatic mood detection from acoustic music data

Music mood describes the inherent emotional meaning of a music clip. It is helpful in music understanding, music search and some music-related applications. In this paper, a hierarchical framework is presented to automate the task of mood detection from acoustic music data, by following some music psychological theories in western cultures. Three feature sets , intensity, timbre and rhythm, are extracted to represent the characteristics of a music clip. Moreover, a mood tracking approach is also presented for a whole piece of music. Experimental evaluations indicate that the proposed algorithms produce satisfactory results.