Melodic segmentation: evaluating the performance of algorithms and musical experts

We review several segmentation algorithms, qualitatively
highlighting their strengths and weaknesses. We also
provide a detailed quantitative evaluation of two existing
approaches, Temperley's Grouper and Cambouropoulos' Local
Boundary Detection Model. In order to facilitate the comparison
of an algorithm's performance with human behavior, we compiled a
corpus of melodic excerpts in different musical styles and
collected individual segmentations from 19 musicians. We then
empirically assessed the algorithms' performance by observing
how well they can predict both the musicians' segmentations and
data taken from the Essen folk song collection.