Date of Award
2009
Degree Type
Thesis
Degree Name
Master of Engineering Science
Program
Electrical and Computer Engineering
Supervisor
Dr. Brown, Lyndon J.
Abstract
In this thesis, two applications are presented. The first one is the polyphonie pitch estimation algorithm based on a previous peer’s work. The algorithm is an iterative multiple-step approach. The input signals are first transformed to a time frequency representation using Instantaneous Fourier Decomposition. Then a Com putational Auditory Scene Analysis based method extracts notes from the time frequency representation. The modifications are presented and compared with previ ous work, results on MIDI music are presented and discussed. The other application is the frequency tracking algorithm based on adaptive inter nal model theory, being able to capture the initial part of a monophonie signal with multiple harmonies. By running an adaptive internal model based closed loop system two times: the first time forwards-in-time, and the second time backwards-in-time
with correct initial values for the state variables calculated using the result in the first run, perfect tracking in the whole period of the signals is achieved. Results on synthesized signals are presented and discussed.
Recommended Citation
Wen, Quan, "Polyphonie Pitch Estimation and Modified Frequency Tracking System" (2009). Digitized Theses. 4755.
https://ir.lib.uwo.ca/digitizedtheses/4755