Date of Award
Master of Engineering Science
Electrical and Computer Engineering
Dr. Vijay Parsa
In this thesis, the application of a subband adaptive model to characterize compression behaviour of five digital hearing aids is investigated. Using a signal-to-error ratio metric, modeling performance is determined by varying the number of analysis bands in the subband structure as well as consideration of three adaptive algorithms. The normalized least mean-squares (NLMS), the affine projection algorithm (APA), and the recursive least-squares (RLS) algorithms are employed using a range of parameters to determine the impact on modeling performance. Using the subband adaptive model to estimate the time-varying frequency response of each hearing aid allows the Perceptual Evaluation of Speech Quality (PESQ) mean-opinion score (MOS) to be computed. The PESQ MOS facilitates an estimation of a subjective assessment of speech quality using an objective score. Initial results suggest the PESQ MOS score is able to differentiate speech processed by hearing aids allowing them to be ranked accordingly. Further work is required to obtain subjective assessments of the processed speech signals and determine if possible correlations exist.
Wirtzfeld, Michael Roy, "Subband Adaptive Modeling of Digital Hearing Aids" (2006). Digitized Theses. 3248.