Files in this item

FilesDescriptionFormat

application/pdf

application/pdfGUO-THESIS-2019.pdf (2MB)Restricted to U of Illinois
(no description provided)PDF

Description

Title:Masking effects of perceptual cues on hearing-impaired ears for speech perception
Author(s):Guo, Siyan
Advisor(s):Allen, Jont
Department / Program:Electrical & Computer Eng
Discipline:Electrical & Computer Engr
Degree Granting Institution:University of Illinois at Urbana-Champaign
Degree:M.S.
Genre:Thesis
Subject(s):Hearing aids fitting
Speech perception
Abstract:The thesis discusses a strategy for evaluating how the hearing-impaired (HI) ear decodes speech sounds by analyzing the impact of manipulating the perceptual speech cues (Li and Allen, 2011), followed by a confusion matrix (CM) error pattern analysis. This study analyzes subjects' responses based on consonant errors. Following this analysis, key insights are provided into the HI listening strategy. Two experiments on consonant-vowel identification were conducted on five normal-hearing (NH) and ten HI subjects. Subjects' performance includes different primary cue intensities (i.e., signal-to-noise ratio) with and without conflicting cues (Kapoor and Allen, 2012). A classification scheme is then introduced to quantify HI ears based on both their error patterns and hearing loss profiles. The idiosyncratic nature of the HI ears is then studied and further quantified using an entropy metric. This analysis demonstrates that: (1) HI ears are using the same primary cue as NH ears, and (2) some HI ears also use conflicting cues for consonant identification. These conclusions are unique to this study (and the unpublished PhD thesis of Cole (2017)). This classification method should be clinically useful in predicting success with the HI aided condition, as well as in improving the insertion gain fitting procedure.
Issue Date:2019-03-27
Type:Text
URI:http://hdl.handle.net/2142/104981
Rights Information:Copyright 2019 Siyan Guo
Date Available in IDEALS:2019-08-23
Date Deposited:2019-05


This item appears in the following Collection(s)

Item Statistics