Identify the Swaras of Carnatic Music using Google’s Machine Learning Model

I always tried to analyze Indian classical ragas and try to figure out the nuances between them. In 2005, I have written a book titled “RagaChintamani” covering the classical aspects of Tamil Film Music.

https://www.amazon.co.uk/Chintamani-Guide-Carnatic-Ragas-Through/dp/B071JVQT3P

I also have great fascination towards automating the raga identification in Indian classical music… Many academic institutions have worked on this project. But since I don’t have extensive coding knowledge, I could not pursue on this. But now with the advent of Machine Learning tools, I wanted to explore on this subject.

Today as a first step, I have tried to explore Google’s Teachable Machine learning model. I tried to come out with a Machine Learning model that can detect the Sapthaswaras of carnatic music.

https://teachablemachine.withgoogle.com/

I tried to teach the system the basic Sapthaswaras of Carnatic Music…. I set A# as my shruthi. Within few minutes and few data-sets, the system got trained…

https://en.wikipedia.org/wiki/Shruti_(music)

Now the model is available for testing… Here is the link for the same…

https://teachablemachine.withgoogle.com/models/XCd3joCaN

You can keep A# as your shruthi and start singing the sapthaswaras one by one (preferably one at a time) and the system is supposed to detect the swaras… It worked reasonably well for me, since the trained voice was from me. You may try testing this model and share your feedback.

This is my first amateurish attempt in this space.I am still working on the ML model to fine-tune it. If this becomes successful, I may try to use the AI/ML system to detect various ragas.

--

--

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store