Matrix-MCE based fuzzy neural network for speech recognition

Author

Gin-Der Wu ; Zhen-Wei Zhu

Author_Institution

Dept. of Electr. Eng., Nat. Chi Nan Univ., Puli, Taiwan

fYear

2014

fDate

18-20 June 2014

Firstpage

546

Lastpage

550

Abstract

Matrix-MCE (MMCE) based fuzzy neural network (FNN) for speech recognition is proposed in this paper. The environment noises usually degrade the performance of speech recognition. To reduce the effect of noises, MMCE is applied to minimize the classification error of two-dimension-cepstrum (TDC). Then the template matching employs FNN. To evaluate the performance, the speech data used for our experiments are a set of isolated Mandarin digits. Experimental results indicate that MMCE-based FNN works better than the other methods.

Keywords

fuzzy neural nets; matrix algebra; pattern classification; pattern matching; speech recognition; FNN; MMCE-based FNN; TDC; environment noises; isolated Mandarin digits; matrix-MCE based fuzzy neural network; minimum classification error; speech data; speech recognition performance; template matching; two-dimensioncepstrum; Fuzzy neural networks; Noise; Noise measurement; Principal component analysis; Robustness; Speech; Speech recognition; fuzzy neural network; speech recognition; two-dimension-cepstrum;

fLanguage

English

Publisher

ieee

Conference_Titel

Control & Automation (ICCA), 11th IEEE International Conference on

Conference_Location

Taichung

Type

conf

DOI

10.1109/ICCA.2014.6870978

Filename

6870978

Link To Document

https://search.isc.ac/dl/search/defaultta.aspx?DTC=49&DC=1775364