This research is directed toward improved speech reception for users of hearing aids through a program of basic and applied research. Our basic research involves a study of the limitations imposed by characteristics of the hearing impairment. Our applied research involves a study of promising signal processing approaches for hearing aids. Research on the effects of hearing impairments focuses on the question of to what extent the effects of the hearing loss on speech perception can be accounted for by reduced audibility. This research will address the role of audibility in the speech-reception performance of listeners with severe-to- profound hearing impairments. We plan to address this question by determining the degree to which functional simulations of the effects of hearing loss that allow s age-matched listeners with normal hearing to experience the loss of sensitivity characteristic of sensorineural hearing impairment. Our basic approach towards hearing-loss simulation employs multiband amplitude expansion combined with broadband noise to achieve frequency-dependent threshold elevation and loudness recruitment. Listeners who are well modelled by these functional simulations are expected to benefit as well from amplitude compression as the listeners with simulated hearing loss. For listeners whose losses are not are not well modelled by functional simulations, we will explore the role of supra-threshold spectral and temporal processing deficits. We will also study the ability of listeners with sensorineural impairments to integrate the cues use to perceive speech across frequency. This work will involve both new experiments and a rigorous model-based analysis of the extent to which deficits in the ability to integrate speech cues exist. The advantage of model basedanalysis is that issues of within-band processing (including loss of audibility as well as supra-threshold factors) can be separated from issues of cross-band integration. Attempts to develop improved signal processing techniques for hearing aids involve the study of vocoder-based frequency lowering for listeners with severe-to-profound high-frequency losses. We will study schemes that incorporate relatively mild amounts of frequency lowering customized for each hearing-impaired listener, combined with amplitude compression, needed by listeners with severe-to-profound hearing losses and recruitment.
Showing the most recent 10 out of 46 publications
