Segmentation of Carnatic music items using KL2, GMM and CFB energy feature

Every Carnatic music concert is made up of many musical items. Every musical item has a lyrical composition (kriti) which can be optionally preceded by an ālāpanā segment. The duration of the ālāpanā along with the rāgā in which the ālāpanā has been rendered is a strong indication of an artist's creativity and musical knowledge. Hence automatic segmentation of an item to extract the ālāpanā segment is of great value in qualitative assessment of a concert. Segmenting a musical item into ālāpanā and kriti has applications in musical retrieval. To find the boundary between ālāpanā and kriti, KL2 distance on Cent Filterbank Energy feature is used that locates change in timbre property. A GMM is used to verify the boundary. To further improve the accuracy of segmentation, rules based on musical domain knowledge are automatically applied. Using this approach a frame-level accuracy of 91.34% was obtained.