Automatic Melakarta Raaga Identification Syste: Carnatic Music

Abstract

Abstract-It is through experience one could as certain that the classifier in the arsenal or machine learning technique is the Nearest Neighbour Classifier. Automatic melakarta raaga identification system is achieved by identifying the nearest neighbours to a query example and using those neighbours to determine the class of the query. This approach to classification is of particular importance today because issues of poor run-time performance are not such a problem these days with the computational power that is available. This paper presents an overview of techniques for Nearest Neighbour classification focusing on; mechanisms for finding distance between neighbours using Cosine Distance, Earth Movers Distance and formulas are used to identify nearest neighbours, algorithm for classification in training and testing for identifying Melakarta raagas in Carnatic music. From the derived results it is concluded that Earth Movers Distance is producing better results than Cosine Distance measure.

Automatic Instrumental Raaga ? A Minute Observation to Find Out Discrete System for Carnatic Music

“… Chronologically, they are sampurna raaga -they contain all seven swaras of the octave in both upword and downword scale [1][2].  They are also known as krama sampurna raaga -that is the sequence is strictly ascending and descending in the scales, without any jumps or hap harard notes [9].…”

“…These features are given to (i) GMM (Gaussian Mixture Model) [1] (ii) HMM (Hidden Markov Model) [2] (iii) KNN (K-Nearest Neighbor) with Cosine distance (iv) KNN with Earth Mover Distance (EMD) [3][4] for training and testing the system. …”

“…The process of identifying the sequence of notes i.e. the swara script is called as Note Transcription [18,2,1]. For Indian classical music, Note Transcription process itself is a very challenging task Music is represented as signal, so Note Transcription involves signal Processing techniques so that it could attract the attention of the audience.…”

“…To simplify, the type of the audio file taken as input is monophonic song with a single sound-source. System can be enhanced further for recorded Polyphonic music, by applying multi-pitch detection techniques described in [9,18,2]. After listing all the frequency segments in the audio file, each such segment has to be converted into its corresponding swara.…”

Data Mining Techniques for Automatic Recognition of Carnatic Raga Swaram Notes

“…The precision values varied from 26% to 72% for the different feature classes. More work on genre classification was reported by Norowi et.al, [9] [10] presented an overview of techniques for Nearest Neighbor classification that placed focus on mechanisms for finding distance between neighbors using Cosine Distance, Earth Movers Distance and formulae were used to identify nearest neighbors and algorithms for classification in training and testing in order to identify Melakartha raga in Carnatic music. From the derived results it was concluded that Earth Movers Distance produced better results than Cosine Distance measure.…”