A Robust Distributed Speech Recognition in Mobile Communications

Author

Djamel, Addou ; Boudraa, Bachir

Author_Institution

Speech & Signal Process. Lab., Univ. of Sci. & Technol., Algiers, Algeria

fYear

2011

fDate

6-8 Dec. 2011

Firstpage

479

Lastpage

483

Abstract

In this paper, a new noise front-end is proposed to improve the performance of Distributed Speech Recognition (DSR) system using a combination of conventional Mel-Cepstral Coefficients (MFCC) and Mel-Line Spectral Frequencies (MLSF). These features are adequately transformed and reduced in a multi-stream scheme using Karhunen-Loeve Transform (KLT). We investigate the performance of a new front-end DSR in terms of recognition accuracy in adverse conditions as well as in terms of dimensionality reduction. Our results showed that for highly noisy speech, using the proposed transformation scheme MLSF-KLT leads to a significant improvement in recognition accuracy on Aurora 2 task.

Keywords

Karhunen-Loeve transforms; mobile radio; speech recognition; Karhunen-Loeve transform; Mel-Cepstral coefficients; Mel-Line spectral frequencies; mobile communications; robust distributed speech recognition; Acoustics; Hidden Markov models; Noise measurement; Robustness; Speech; Speech recognition; Vectors; : Distributed Speech Recognition (DSR); KLT.; MFCC; MLSF; front-end;

fLanguage

English

Publisher

ieee

Conference_Titel

Developments in E-systems Engineering (DeSE), 2011

Conference_Location

Dubai

Print_ISBN

978-1-4577-2186-1

Type

conf

DOI

10.1109/DeSE.2011.94

Filename

6149979