ZCPA features for speech recognition

Author

Kacur, Juraj ; Varga, M. ; Rozinaj, Gregor

Author_Institution

Inst. of Telecommun., STU, Bratislava, Slovakia

fYear

2012

fDate

25-27 Oct. 2012

Firstpage

1

Lastpage

4

Abstract

In this article we present implementation, modifications and optimization of zero-crossing peak amplitude (ZCPA) speech feature extraction method into Slovak speech recognition system. ZCPA features are closely mimicking the human auditory system in the time domain, and thus they should be more robust against common noises. Except the basic configuration several modifications have been suggested, implemented and evaluated. Furthermore, optimization of settings on a real system using professional database and MASPER training procedure have been found and compared to classical features presented by MFCC and PLP in different scenarios and noise conditions.

Keywords

feature extraction; natural language processing; speech processing; speech recognition; MASPER training procedure; Slovak language; Slovak speech recognition system; ZCPA features; human auditory system; speech feature extraction; time domain; zero crossing peak amplitude; Filter banks; Histograms; Mel frequency cepstral coefficient; Robustness; Speech; Speech recognition; Training; HMM; MASPER; MFCC; PLP; ZCPA; speech features; speech recognition;

fLanguage

English

Publisher

ieee

Conference_Titel

Telecommunications (BIHTEL), 2012 IX International Symposium on

Conference_Location

Sarajevo

Print_ISBN

978-1-4673-4875-1

Electronic_ISBN

978-1-4673-4874-4

Type

conf

DOI

10.1109/BIHTEL.2012.6412092

Filename

6412092