Further reduced form of Karhunen-Loeve transform for text independent speaker recognition

被引:11
作者
Lung, SY [1 ]
Chen, CCT [1 ]
机构
[1] Natl Sun Yat Sen Univ, Dept Elect Engn, Kaohsiung, Taiwan
关键词
D O I
10.1049/el:19980914
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
A text independent speaker recognition system based on the Karhunen-Loeve transform, derived from the split-and-merge algorithm, is proposed. The split-and-merge algorithm is applied to speaker data compression in the time domain. A set of experiments is conducted which gives a 91% recognition rate for 100 Mandarin speakers.
引用
收藏
页码:1380 / 1382
页数:3
相关论文
共 4 条
[1]   Hard-limited Karhunen-Loeve transform for text independent speaker recognition [J].
Chen, CCT ;
Chen, CT ;
Tsai, CM .
ELECTRONICS LETTERS, 1997, 33 (24) :2014-2016
[2]   Disk distance measure of speaker recognition [J].
Lung, SY .
ELECTRONICS LETTERS, 1997, 33 (20) :1678-1679
[3]   LONG-TERM FEATURE AVERAGING FOR SPEAKER RECOGNITION [J].
MARKEL, JD ;
OSHIKA, BT ;
GRAY, AH .
IEEE TRANSACTIONS ON ACOUSTICS SPEECH AND SIGNAL PROCESSING, 1977, 25 (04) :330-337
[4]  
RUSS JC, 1995, IMAGE PROCESSING HDB, P399