Evaluation of Speaker Identification System using GSMEFR speech Data

Author

Krobba, Ahmed ; Debyeche, Mohamed ; Amrouche, Abederrahmane

Author_Institution

Speech Commun. & Signal Process. Lab., USTHB, Algiers, Algeria

fYear

2010

fDate

23-25 March 2010

Firstpage

1

Lastpage

5

Abstract

This paper investigate the influence of GSMEFR speech Data on the performance of a text independent Speaker Identification System (SIS) based on Gaussian Mixture Models (GMM) classifiers. The performance evaluation due to the use of the GSMEFR speech Data, obtained by passing the local ARADIGIT database through the GSM coder/decoder. The recognition evaluation was also conducted using original ARADIGIT sampled at 16 KHz and its 8 KHz downsampled version. The ARADIGIT database consists of 60 speakers (31 men and 29 women) pronouncing the ten Arabic digits three times each. Different experiments were carried to measure the degradation introduced by different aspects of the simulated codec.

Keywords

cellular radio; codecs; speech recognition; ARADIGIT database; GSM coder/decoder; GSMEFR speech data; Gaussian mixture models classifiers; frequency 16 kHz; frequency 8 kHz; speaker identification system; Databases; Decoding; GSM; Mathematical model; Signal design; Speech analysis; Speech codecs; Speech coding; Speech enhancement; Wireless communication; GSMEFR Speech Data; Gaussian Mixture Models; Speaker Identification;

fLanguage

English

Publisher

ieee

Conference_Titel

Design and Technology of Integrated Systems in Nanoscale Era (DTIS), 2010 5th International Conference on

Conference_Location

Hammamet

Print_ISBN

978-1-4244-6338-1

Type

conf

DOI

10.1109/DTIS.2010.5487589

Filename

5487589