Faraaz Nadeem

Faraaz  Nadeem
MIT EECS | CS+HASS Undergraduate Research and Innovation Scholar
Advisors: Leslie P. Kaelbling
Eran Egozy
Department: EECS
Areas of Research: CS+HASS
Years: 2018-2019
Research Project Title:

Automatic Music Transcription with Multiple Instruments

abstract:The goal of my SuperUROP project is to be able to automatically transcribe musical performances with multiple instruments using computers. In general, the task of music transcription is about listening to a piece of music and writing down all of its musical elements using some sort of notation. The current state-of-the-art models use deep learning to transcribe live piano performances to sheet music. I am working towards extending this research to be able to work with any combination of instruments, instead of just the piano.

"I'm concentrating in music, and I'm very involved in the dance community at MIT. So I'm especially excited about this SuperUROP project because I have the opportunity to help computers develop an understanding of the complex layers that underlie the overall sound of a song in the same way humans compose music in parts or dance to different musical elements!"