Music Classification using Bispectrum and Mel-Frequency Cepstrum Coefficients and artificial neural network classifier (ANN) Techniques

Chandanpreet Kaur
2015 Chandanpreet Kaur International Journal of Electronics, Electrical and Computational System IJEECS   unpublished
In this paper we investigate information pertaining to the pre-processing techniques in music for automatically identifying ragas. We give brief explanation on music structure and traditional techniques for music analysis.This paper used higher order spectra to obtain phase non-linearity and classificationof individual ragas. We compare and present the performances ofthe bispectrum and mel-frequency cepstrum coefficients (MFCC) by using artificial neural network classifier (ANN). 1.
more » ... Music analysis is an attempt to answer the question, how does music work. Music analysis is a vibrant and rapidly evolving field of research, which can enrich the wider signal processing community with exciting applications and new problems. Music is most sophisticated and precisely constructed form of any sound signal, analysis is a practical tool which can help in the understanding of how tonal (and other forms of music) are constructed. For extracting the information of relevance to listeners, requires the different kind of specialized methods of music specific characteristics including harmony, rhythm, pitches, and instrumentation [1]. The purpose of musical analysis is to examine the surface voice leading and then segment it in layers in order to crack the elemental structure of the music. Understanding these elemental structures can help performers in several ways, for instance, with the phrasing of music. By understanding how the underlying phrase structures work, how complete and incomplete structures can be joined, it is possible to understand how feelings of expectation and realization, tension and resolution can be created in music. Understanding how this process works can help with the phrasing and expression of the music. Music has a language that people all over the world can understand at a subconscious level. Large numbers of people round the world listen to music that is tonal or utilises tonal harmony and tonal structures. Tonal harmony has enlarged and weighted music all around the world. So it is worthwhile trying to understand how this enlarged language works and how it varies. It is worth trying to understand the syntax of tonal music and the structure of individual pieces of music simply because listening to it is important to so much of humanity.