ABSTRACT
A scheme has been developed to compensate partly for reduced frequency selectivity. The scheme uses a mathematical optimization procedure to enhance spectral contrasts in order to produce, as nearly as possible, a normal excitation pattern in an impaired ear with reduced frequency selectivity. The enhancement compensated for only a small portion of the loss of intelligibility due to smearing. Moderate degrees of reduced frequency selectivity had little effect on the intelligibility of speech stimuli presented in quiet, but caused substantial reduction in intelligibility when the stimuli were mixed with continous speech-shaped noise, even at speech-to-noise ratios that posed few difficulties for normal listeners. The goal of the signal-processing algorithm is to produce, as nearly as possible, a normal excitation pattern in an auditory system with reduced frequency selectivity. Spectral contrast enhancement algorithms have been developed in our laboratory and, in two previous studies, achieved modest success in improving speech intelligibility in noise for hearing-impaired listeners.
