Saanvi Bhargava

How can we develop an algorithm to provide cumulative feedback in vocal music education?

Started Apr. 15, 2023

Abstract or project description

Scalable music education requires giving fast feedback on student audio performances. Current feedback mechanisms are simply manual and given by teachers, rendering them subjective and therefore mostly ineffective. Technological feedback mechanisms evaluate whether a student is correct on a single note rather than the entire music piece, not providing cumulative or holistic feedback. An algorithm is presented for automatically grading vocal music recordings cumulatively on pitch and rhythm given an expected piece of sheet music. The algorithm results in a numerical grade given to the student and overall feedback on improvements the student can make. The algorithm is tested for accuracy on a dataset of corresponding audio recordings, notated music, and a list of mistakes made in the recording. The results show that human feedback is often subjective and inaccurate and that a grading model can be created for vocal music education. My algorithm for scoring the recordings in 4 different categories performed well, reaching an accuracy of 61.4%.