A NEURO-FUZZY APPROACH TO SPEECH RECOGNITION WITHOUT TIME ALIGNMENT

Several successful approaches to speech recognition have been proposed . Most of them involve time alignment which requires substantial compu tation and considerable memory storage. In this paper, we present a ne uro-fuzzy approach to speech recognition without time alignment. This approach is a powerful method for selecting reference templates; there fore, considerable memory storage is alleviated. In addition, it great ly reduces substantial computation in the matching process because it obviates time alignment. Base on this approach, a Mandarin speech reco gnition system without time alignment is implemented. Two databases we re utilized for verifying its performance. An encouraging experimental result confirms the effectiveness of the proposed neuro-fuzzy approac h to speech recognition without time alignment. (C) 1998 Elsevier Scie nce B.V. All rights reserved.