An analysis of the effect of basilar membrane nonlinearities on noise suppression.

Computational models of the peripheral auditory system have largely modeled basilar membrane (BM) mechanics as a linear filter-bank-like entity. Recent mathematical work on the nature of auditory system noise suppression allows us to analyze and argue for the incorporation of BM nonlinearities into these models. This analysis shows that vowel perception improves with increasing presence of BM nonlinearities whereas consonant perception degrades with increasing influence of BM nonlinearities. Experimental results on tones and real speech corrupted by noise corroborate the analysis as well as suggest a novel approach to speech processing.