2D CLAFIC subspace technique in probabilistic speaker verification

L. Y. Chong, Andrew B.J. Teoh, S. E. Khor

Research output: Chapter in Book/Report/Conference proceedingConference contribution

2 Citations (Scopus)

Abstract

User-specific subspace method in speaker recognition, such as CLAFIC is used to construct the individual subspace which represents the distinct spectral characteristic for each speaker. It concatenates the speech matrices side by side before forming the correlation matrix. In this paper, we proposed a new method, coined as Probabilistic 2D CLAFIC which applied a straightforward two-dimensional (2D) speech matrix for feature dimension reduction to improve the discrimination and reduce the computation complexity. Gaussian Mixture Model (GMM) is used instead of norm of the projected feature in conventional CLAFIC to boost up the performance. Experimental results showed that our proposed method attained an encouraging performance with the best Equal Error Rate (EER) of 0.56% with full feature dimension, and EER of 2.98% with 50% reduction of feature dimension compare to baseline GMM, 5.36%.

Original languageEnglish
Title of host publication2007 IEEE Workshop on Automatic Identification Advanced Technologies - Proceedings
Pages204-208
Number of pages5
DOIs
Publication statusPublished - 2007 Oct 2
Event2007 IEEE Workshop on Automatic Identification Advanced Technologies, AUTOID 2007 - Alghero, Italy
Duration: 2007 Jun 72007 Jun 8

Publication series

Name2007 IEEE Workshop on Automatic Identification Advanced Technologies - Proceedings

Other

Other2007 IEEE Workshop on Automatic Identification Advanced Technologies, AUTOID 2007
CountryItaly
CityAlghero
Period07/6/707/6/8

All Science Journal Classification (ASJC) codes

  • Computer Science(all)

Cite this

Chong, L. Y., Teoh, A. B. J., & Khor, S. E. (2007). 2D CLAFIC subspace technique in probabilistic speaker verification. In 2007 IEEE Workshop on Automatic Identification Advanced Technologies - Proceedings (pp. 204-208). [4263241] (2007 IEEE Workshop on Automatic Identification Advanced Technologies - Proceedings). https://doi.org/10.1109/AUTOID.2007.380620