9th Annual Conference of the International Speech Communication Association

Brisbane, Australia
September 22-26, 2008

Consonant Discrimination of Degraded Speech Using an Efferent-Inspired Closed-Loop Cochlear Model

David P. Messing (1), Lorraine Delhorne (1), Ed Bruckert (2), Louis D. Braida (1), Oded Ghitza (2)

(1) MIT, USA; (2) Sensimetrics Corporation, USA

We present a model of auditory speech processing capable of predicting consonant confusions by normal hearing listeners, based on a phenomenological model of the Medial Olivocochlear efferent pathway. We then use this model to predict human error patterns of initial consonants in consonant-vowel-consonant words. In the process we demonstrate its potential for speech identification in noise. Our results produced performance that was robust to varying levels of additive noise and which was similar to human performance in discrimination of synthetically spoken consonants.

Full Paper

Bibliographic reference.  Messing, David P. / Delhorne, Lorraine / Bruckert, Ed / Braida, Louis D. / Ghitza, Oded (2008): "Consonant discrimination of degraded speech using an efferent-inspired closed-loop cochlear model", In INTERSPEECH-2008, 1052-1055.