Carnatic Music is based on a particular genre of Indian Music belonging to South India. Similar to Hindustani Music, Carnatic Music is also based on the oral tradition. In this music, there is hardly any music that is annotated or archived. The purpose of the research is to develop signal processing and machine learning algorithms that can be used to characterise and organise Carnatic Music. Five different tasks are addressed:
1. Tonic identification (solved)
2. Segmentation of Carnatic music concerts using applauses.
3. Segmentation of a song in Carnatic music using the pallavi line.
4. Motivic analysis of determining typical motifs of raga.
5. Onset detection and stroke recognition for percussion instruments
6. Modified Group Delay ­Gram Based Music Source Separation Using Deep Recurrent Neural Networks
Summary Report