Efficient window for monolingual and crosslingual speaker identification using MFCC

Author

Nagaraja, B.G. ; Jayanna, H.S.

Author_Institution

Dept. of Electron. & Commun. Eng., GM Inst. of Technol., Davangere, India

fYear

2013

fDate

19-21 Dec. 2013

Firstpage

1

Lastpage

4

Abstract

In this paper an experimental evaluation of the various windowing techniques using mel-frequency cepstral coefficient (MFCC) for monolingual and crosslingual speaker identification is demonstrated. The set of windows presented here allows a tradeoff between main lobe bandwidth and side lobe ripple decay. The speaker identification study is conducted using randomly selected 50 speakers from IITG Multi-variability speaker recognition (IITG-MV) database, MFCC feature and Gaussian mixture model (GMM)-universal background model (UBM) classifier. Speaker identification system based on various windowing techniques shown to have considerably improved performance over baseline Hamming window technique.

Keywords

Gaussian processes; mixture models; signal classification; speaker recognition; GMM; Gaussian mixture model-universal background model classifier; IITG multivariability speaker recognition database; IITG-MV; MFCC feature; UBM; baseline Hamming window technique; crosslingual speaker identification; main lobe bandwidth; mel-frequency cepstral coefficient; monolingual speaker identification; side lobe ripple decay; speaker identification study; windowing techniques; Communication systems; Databases; Frequency-domain analysis; Mel frequency cepstral coefficient; Speaker recognition; Speech; Speech recognition; MFCC; Speaker identification; crosslingual; monolingual; window;

fLanguage

English

Publisher

ieee

Conference_Titel

Advanced Computing and Communication Systems (ICACCS), 2013 International Conference on

Conference_Location

Coimbatore

Type

conf

DOI

10.1109/ICACCS.2013.6938702

Filename

6938702