DSP-based fuzzy neural networks and its application in speech recognition

No Thumbnail Available

Date

1999-10-15

Authors

S.-C. Chen
C.-C. Hsu
W.-Y. Wang

Journal Title

Journal ISSN

Volume Title

Publisher

Abstract

A fuzzy-neural network needs to be trained through a learning process, so that suitable membership functions and weightings can be obtained. However, most neural networks are only simulated by computer software, which are not practical for real applications. It is therefore our objective to design an integrated circuit system based on a DSP processor with powerful arithmetical capabilities and fast data processing, and relevant peripheral devices to implement the fuzzy neural network. In terms of implementation cost and feasibility for practical applications, this DSP-based fuzzy neural network will be more practical and usable. Finally, a prospective application of the DSP processor-based fuzzy neural network to recognize speech from a non-designated person is proposed

Description

Keywords

Citation

Collections