HANİLÇİ, Cemal Uludağ University - Faculty of Engineering and Architecture - Dept of Electronic Engineering, Turkey , ERTAŞ, Figen Uludağ University - Faculty of Engineering and Architecture - Dept of Electronic Engineering, Turkey

EFFECTS OF BACKGROUND DATA DURATION ON SPEAKER VERIFICATION PERFORMANCE

34743

Gaussian mixture models with universal background model (GMM-UBM) and vector quantization with universal background model (VQ-UBM) are the two well-known classifiers used for speaker verification. Generally, UBM is trained with many hours of speech from a large pool of different speakers. In this study, we analyze the effect of data duration used to train UBM on text-independent speaker verification performance using GMM-UBM and VQ-UBM modeling techniques. Experiments carried out NIST 2002 speaker recognition evaluation (SRE) corpus show that background data duration to train UBM has small impact on recognition performance for GMM-UBM and VQ-UBM classifiers.

111

Speaker verification , Gaussian mixture model , Vector Quantization , Universal background model

Uludağ University Journal of The Faculty of Engineering

119