TY - GEN
T1 - 2D CLAFIC subspace technique in probabilistic speaker verification
AU - Chong, L. Y.
AU - Teoh, Andrew B.J.
AU - Khor, S. E.
PY - 2007
Y1 - 2007
N2 - 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%.
AB - 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%.
UR - http://www.scopus.com/inward/record.url?scp=34748926052&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=34748926052&partnerID=8YFLogxK
U2 - 10.1109/AUTOID.2007.380620
DO - 10.1109/AUTOID.2007.380620
M3 - Conference contribution
AN - SCOPUS:34748926052
SN - 1424412994
SN - 9781424412990
T3 - 2007 IEEE Workshop on Automatic Identification Advanced Technologies - Proceedings
SP - 204
EP - 208
BT - 2007 IEEE Workshop on Automatic Identification Advanced Technologies - Proceedings
T2 - 2007 IEEE Workshop on Automatic Identification Advanced Technologies, AUTOID 2007
Y2 - 7 June 2007 through 8 June 2007
ER -