Genre Classification of Indian Tamil Music using Mel-Frequency Cepstral Coefficients

Abstract

Musical genres are categorical labels created by humans to characterize pieces of music that are related by common characteristics such as instrumentation, rhythmic structure, and harmonic content of the music. This paper presents the Automatic Genre Classification of Indian Tamil Music using Timbral features and Mel-Frequency Cepstral Coefficient features. The classifier model for the proposed system has been built using the K-Nearest Neighbours and Support Vector Machine classifiers. The performance of various features extracted from music excerpts has been analyzed, to identify the appropriate feature descriptors for the two major genres of Indian Tamil music, namely Classical music and Folk music. The results have shown that the feature combination of Spectral Roll off, Spectral Flux, Spectral Skewness and Spectral Kurtosis, combined with Mel-Frequency Cepstral Coefficient features yields a classification accuracy of 84.21 % .