Title :
Indian instrumental music: Raga analysis and classification
Author :
Yogesh H. Dandawate;Prabha Kumari;Anagha Bidkar
Author_Institution :
VIIT, Pune, Maharashtra, India
Abstract :
Raga played by Indian instrument is the actually soul of the Indian classical music. Indian classical music is famous around all over the world for its particular structure and well soundness. Our work is related to analyze and classify the instrumental music according to their features. This will help to non-professional and music learner for understanding and acquire knowledge about the music using the system intelligence. There are various features for analysis of music but our approach is towards the spectral and temporal features. For extraction of feature we take spectrum, chromagram, centroid, lower energy, roll off, histogram etc. At very first we just collect clips of ragas and find out the spectral and temporal features. These features show the better result. We are using four ragas namely:- Bhairav, Bhairavi, Todi and Yaman. For classification we use different types of classifier just like KNN classifier and SVM classifier they gives approximate 87% and 92% accuracy respectively.
Keywords :
"Mathematical model","Instruments","Feature extraction","Support vector machines","Histograms","Frequency measurement","Propulsion"
Conference_Titel :
Next Generation Computing Technologies (NGCT), 2015 1st International Conference on
DOI :
10.1109/NGCT.2015.7375216