Synthesis and performance analysis of a Recurrent Fuzzy Multilayer Perceptron for speech recognition

Author

Gupta, Akshay

Author_Institution

Dept. of IT, Bharati Vidyapeeth´´s Coll. of Eng., New Delhi, India

fYear

2010

fDate

13-14 Dec. 2010

Firstpage

22

Lastpage

26

Abstract

A novel speech recognition method has been proposed which combines the capabilities of a Recurrent Fuzzy Multilayer Perceptron (MLP) to the existing Mel Frequency Cepstral Coefficients (MFCC) model, synthesized using JAVA. Performance analysis of the proposed recurrent fuzzy MLP relative to a speech recognition system has been shown using MATLAB. Owing to its short-term memory effect in addition to inherent neuro-fuzzy model advantages, the simulation results obtained were significantly better than offered by a crisp neural network.

Keywords

Java; cepstral analysis; multilayer perceptrons; recurrent neural nets; speech recognition; JAVA; MATLAB; Mel frequency cepstral coefficients model; recurrent fuzzy multilayer perceptron; speech recognition method; Analytical models; Artificial neural networks; Frequency synthesizers; Hidden Markov models; Mel frequency cepstral coefficient; Nerve fibers; Speech recognition; artificial neural networks; fuzzy logic; multilayer perceptron; recurrent neural networks; speech recognition;

fLanguage

English

Publisher

ieee

Conference_Titel

Methods and Models in Computer Science (ICM2CS), 2010 International Conference on

Conference_Location

New Delhi

Print_ISBN

978-1-4244-9701-0

Type

conf

DOI

10.1109/ICM2CS.2010.5706713

Filename

5706713