Second International Conference on Spoken Language Processing (ICSLP'92)

Banff, Alberta, Canada
October 13-16, 1992

Auditory Model Based Speech Processing

Yuqing Gao (1,2), Taiyi Huang (1), Shaoyan Chen (1), Jean-Paul Haton (2)

(1) National Laboratory of Pattern Recognition, Institute of Automation , Chinese Academy of Sciences, Beijing, China
(2) CRIN/CNRS, Vandoeuvre-les-Nancy, France

This paper describes a, computational auditory model that be designed to reflect the processing steps and activities occurring in the human auditory systemmm. The model consists of the periphery auditory model and feedback control model from central nervous system to the cochlear. The feedback model is developed based on the efferent-induced effect observed in physiological experiments. The auditory model with and without feedback circuit are compared with the LPC analysis method using clean speech and noisy speech to stimulate the model in order to evaluate the speech processing capability of the model. The preliminary speech recognition results shows that the auditory model is effective for speech processing.

Full Paper

Bibliographic reference.  Gao, Yuqing / Huang, Taiyi / Chen, Shaoyan / Haton, Jean-Paul (1992): "Auditory model based speech processing", In ICSLP-1992, 73-76.