Classifying polyphony music based on Markov model

Abstract

In this investigation we propose a novel approach for classifying polyphonic melodies. Our main idea comes from Probability Stochastic Processes using Markov models where the characteristic features of polyphonic melodies are extracted from each bar. The similarity among harmonies can be considered by means of the features. We show the effectiveness and the usefulness of the approach by experimental results. © Springer-Verlag Berlin Heidelberg 2006.