590 Course Seminar - Alara Ece Şensoy - Chord Recognition with ML

DEPARTMENT OF COMPUTER ENGINEERING

CSE 590 COURSE SEMINAR

Chord Recognition with ML

Alara Ece Şensoy

Date & Time: Monday, January 27, 2025, 16:00

Place: https://meet.google.com/ovx-cpww-ath

All interested are cordially invited.

ABSTRACT

Without having absolute pitch, it is near impossible to immediately recognize chords by ear and play them on an instrument. Since there are no readily available sheets for every musical piece out there, recognizing chords using machine learning comes in handy. With this seminar, the aim is to explain the in-use, out-of-use (for comparison), and hybrid models to show which ones could be benefitted from in order to get a better consistency at recognizing accurate chords. Currently, models, such as HMM, are not really used for sound processing all that much. However, when hybrid models of these models’ combinations, like CNN-HMM, are used, they yield better results. There are also newer models to be mentioned, which have not yet been integrated enough into the academy judging from various papers, that will be touched upon.

BIOGRAPHY

Alara Ece Şensoy is an M.Sc. student in the Computer Engineering Program of Yeditepe University. She also had her Bachelor’s Degree in Computer Engineering from the same department in 2023. Having gone to Istanbul University State Conservatory and playing the double bass for about a year while attending Robert College, she has always been interested in the field of Music and musical instruments. Currently, while carrying out her duties as a T.A., she does research on Sound and Image Processing coupled with Machine Learning as a member of the CNVP Lab.