HIERARCHICAL MIXTURE CLUSTERING AND ITS APPLICATION TO GMM BASED TEXT INDEPENDENT SPEAKER IDENTIFICATION

被引：0

作者：

Saeidi, R. ^{[1
]}

Mohammadi, H. R. Sadegh ^{[1
]}

Ganchev, T. ^{[3
]}

Rodman, R. D. ^{[2
]}

机构：

[1] Iranian Res Inst Elect Engn, Tehran, Iran

[2] North Carolina State Univ Raleigh, Dept Comp Sci, Raleigh, NC USA

[3] Univ Patras, Wire Commun Lab, Patras 26500, Greece

来源：

2008 INTERNATIONAL SYMPOSIUM ON TELECOMMUNICATIONS, VOLS 1 AND 2 | 2008年

关键词：

Speaker identification; mixture clustering; GMM; speed-up; VERIFICATION;

D O I：

10.1109/ISTEL.2008.4651403

中图分类号：

TM [电工技术]; TN [电子技术、通信技术];

学科分类号：

0808 ; 0809 ;

摘要：

In this paper, we propose a hierarchical mixture clustering method and investigate its application for complexity reduction of a GMM based speaker identification system. We show that by using GMM-HMC one can cluster speakers more accurately than that of a sorted GMM with the same acceleration rate. The system was tested on a universal background model-Gaussian mixture model with KL-divergence as the distance measure. While the proposed system's performance is slightly inferior to the baseline system, its comparatively smaller computational load provides the potential to develop systems with higher performance.

引用

页码：770 / +

页数：2

共 13 条

[1]

AUCKENTHALER R, 2001, P SPEAK OD SPEAK REC

[2]

Chan Arthur., 2004, Proceedings of INTERSPEECH 2004, P689

[3]

FORTUNA J, P INTERSPEECH 2005

[4] Real-time speaker identification and verification [J].