Psychoacoustical criteria and models for individual optimization of signal enhancement in hearing aids
Summary:
Some signal enhancement algorithms for hearing aids, especially single-channel noise suppression and fast multi-band compression, must be evaluated and optimized using signal-quality measures that also account for the end-user's auditory capacity. This project will develop physical (objective) methods to predict speech intelligibility and the most important aspects of signal quality, as perceived by listeners with impaired hearing, based on physical measurements on the original and the processed signal, and on psycho-acoustical measurements of the listener's auditory acuity. The resulting measures will be used to evaluate existing signal enhancement methods, and to create a new theoretical foundation and practical procedures for the tuning of algorithms to individual hearing impairment.
References:
Derleth, R.P., Dau, T., and Kollmeier, B. (2001). Modeling temporal and compressive properties of the normal and impaired auditory system. Hearing Research, 159:132–149.
Leijon, A. (2002). Estimation of auditory information transmission capacity using a hidden Markov model of speech stimuli. Acustica - Acta Acustica, 88:3, pp. 423-32.
Stadler, S., Leijon, A., and Hagerman, B. (2007). An information theoretic approach to estimate speech intelligibility for normal and impaired hearing. In Interspeech 07, Antwerpen, BE.
