Basic Information
- Project Title: Audio Source Separation, Melody Extraction
- Name: Sharvaree Sinkar
- Project: SRE
- Semester(s): 7,8
- Guide: Prof. Preeti Rao
Abstract
Did two projects under Preeti Ma’am
- Audio Source Separation: Implemented NMF based method for source separation on Indian Classical music dataset having harmonic and percussion components. Performed ‘Piano Chord Detection’ and ‘Notes Transcription’ as the part of project.
- Melody Extraction: Did literature review of existing salience based, source separation based and ML based techniques on melody extraction. Implemented Melodia algorithm on Indian Classical Dataset
Combination of both of these projects is something I am working on my DDP.
DDP: Melody Extraction and Pattern Matching(Synchronisation) for Generative Music
Any courses you completed relevant to the project
EE679: Speech Processing
Describe your experience on the project
The project included application of Signal Processing and ML/AI techniques on real life applications, which was something interesting. The project work was both theory and coding heavy. There was a meet scheduled once a week.
Describe your experience with the guide
Preeti Ma’am gave time to time feedbacks which were thorough enough. We met once a week. Sometimes, once in two weeks. In terms of implementation, she is very particular of what code you wrote, she delves into depth and sees if we have understood the concept thoroughly.
Evaluation consisted of a PPT and a report submission
Any advice for anyone considering a project under the same guide? Any other professors working in similar fields?
Be alert to reply to Ma’am’s msgs on Teams or mail. She needs time to time updates