Influence of Features Extraction Methods in Performance of Continuous Speech Recognition for Romanian

Author

Dumitru, C.O. ; Gavat, Inge

Author_Institution

Univ. Politehnica Bucharest, Bucharest

fYear

2007

fDate

27-30 June 2007

Firstpage

288

Lastpage

291

Abstract

This paper describes continuous speech recognition experiments for Romanian language, based on statistical modelling by using hidden Markov models. These experiments are made in order to select the most appropriate features extraction method. The compared methods are cepstral and LPC analysis, in standard and perceptual versions. In our tests the cepstral coefficients perform in the most situations better versus the linear prediction ones, and the perceptual coefficients better than the standard ones.

Keywords

cepstral analysis; feature extraction; hidden Markov models; linear predictive coding; natural language processing; speech processing; speech recognition; LPC analysis; Romanian language; cepstral analysis; continuous speech recognition; features extraction method; hidden Markov models; perceptual coefficients; statistical modelling; Cepstral analysis; Cepstrum; Discrete Fourier transforms; Feature extraction; Hidden Markov models; Linear predictive coding; Mel frequency cepstral coefficient; Signal processing; Speech recognition; Testing; HMM; LPC; MFCC; PLP; WRR; statistical modelling;

fLanguage

English

Publisher

ieee

Conference_Titel

Systems, Signals and Image Processing, 2007 and 6th EURASIP Conference focused on Speech and Image Processing, Multimedia Communications and Services. 14th International Workshop on

Conference_Location

Maribor

Print_ISBN

978-961-248-029-5

Electronic_ISBN

978-961-248-029-5

Type

conf

DOI

10.1109/IWSSIP.2007.4381098

Filename

4381098