Research
I am a PhD student at Music Technology Group (MTG) of University of Pompeu Fabra, Barcelona, Spain. I am working as a part of CompMusic project with Prof. Xavier Serra. My research interests are based around Music Information Retrieval (MIR), which involves techniques from information retrieval, signal processing and machine learning applied to audio music signals.
Currently my focus is towards exploring culture specific computational model for melodic analysis of Indian art music. Some of the specific topics that I am working on within the CompMusic project are:
Currently my focus is towards exploring culture specific computational model for melodic analysis of Indian art music. Some of the specific topics that I am working on within the CompMusic project are:
- Identification of Tonic of the lead artist
- Melodic feature extraction and its optimal representation
- Pattern discovery in melodic sequence (melodic motives)
- Characterization of melodic motives
Projects
Discovery and Characterization of Melodic Motives in Large Audio Music Collection of Indian Art Music (2012-present)
This is the topic of my current research (during PhD) as a part of CompMusic project, under the supervision of Prof. Xavier Serra. I also collaborate with Joan Serrà in several topics in my PhD work.
Melodic motives are musically repeating patterns in the melody, which characterize certain aspects of music. In the case of Indian art music there is a special significance of these melodic motives (also referred as Pakads), as they are the main cues for a listener to identify a rāg. A basic demo of this work can be found here.
Melodic motives are musically repeating patterns in the melody, which characterize certain aspects of music. In the case of Indian art music there is a special significance of these melodic motives (also referred as Pakads), as they are the main cues for a listener to identify a rāg. A basic demo of this work can be found here.
A Tonic Identification Approach for Indian Art Music (2011-12):
In Indian art music the tonic pitch chosen by the lead performer is a fundamental musical (tonal) information. Since, this pitch serves as a reference for the tonal organization in the musical piece, its identification is a crucial first step for a more detailed automatic analyses of melodic aspects. This was my masters project in MTG-UPF, you can download the thesis document from here (MTG).
Rhythm analysis of Indian semi-classical music (2010-2011)
This project focused on estimation of perceptual tempo and meter from audio songs of Indian semi-classical music. The work was supervised by Prof. Preeti Rao, in Digital Audio Processing lab in Indian Institute of Technology Bombay, Mumbai.
You can read more in the rhythm related publications.
You can read more in the rhythm related publications.