Third International Conference on Spoken Language Processing (ICSLP 94)

Yokohama, Japan
September 18-22, 1994

Acoustic Simulation of Auditory Model Based Speech Processor for Cochlear Implant System

Shigeyoshi Kitazawa (1), Kazuyuki Muramoto (1), Juichi Ito (2)

(1) Shizuoka University, Hamamatsu, Japan
(2) Ohtsu Red Cross Hospital, Ohtsu, Japan

We propose a speech processor based on the auditory model to generate stimulus pattern similar to the auditory system. Our model consists of the auditory filter to simulate the tuning curve and the hair-cell model to simulate the non-linear characteristics of the auditory nerve firing. The model consists of 28 separate channels placed equidistant along the basilar membrane. Assuming the implanted receiver stimulator, the stimulation is possible sequentially at the interval greater than 0.8 ms. The fundamental period was obtained from the summation of lower three channels. During every period, four most frequent firing channels were selected to stimulate- We estimated the quality of electric sensation by acoustic simulation. We carried out listening experiment by normal hearing subjects without visual presentation. The subjects have got improvements in perception of voiced stops, voiced fricatives, voiceless fricatives, voiced affricates, and nasals.

Full Paper

Bibliographic reference.  Kitazawa, Shigeyoshi / Muramoto, Kazuyuki / Ito, Juichi (1994): "Acoustic simulation of auditory model based speech processor for cochlear implant system", In ICSLP-1994, 2043-2046.