TEXT-DEPENDENT GMM-JFA SYSTEM FOR PASSWORD BASED SPEAKER VERIFICATION

被引：0

作者：

Novoselov, Sergey ^{[1
]}

Pekhovsky, Timur ^{[1
]}

Shulipa, Andrey ^{[1
]}

Sholokhov, Alexey ^{[1
]}

机构：

[1] Speech Technol Ctr Ltd, St Petersburg, Russia

来源：

2014 IEEE INTERNATIONAL CONFERENCE ON ACOUSTICS, SPEECH AND SIGNAL PROCESSING (ICASSP) | 2014年

关键词：

speaker recognition; NAP; SVM; JFA; UBM; GMM; HMM; supervector;

D O I：

暂无

中图分类号：

O42 [声学];

学科分类号：

070206 ; 082403 ;

摘要：

We propose a new State-GMM-supervector extractor for solving the problem of text-dependent speaker recognition. The proposed scheme for supervector extraction makes it easy to implement a text-dependent JFA system for passphrase verification. We examine the conditions of both a global and a text-prompted passphrase. The experiments conducted on the Wells Fargo Bank speech database show that the proposed method makes it possible to create more accurate statistical models of speech signals and to achieve a 44% relative reduction of EER compared to the best state-of-the-art systems of text-dependent verification for a text-prompted passphrase.

引用

页数：5

共 50 条

[41] Robust distant speaker recognition based on position-dependent CMN by combining speaker-specific GMM with speaker-adapted HMM
Wang, Longbiao
Kitaoka, Norihide
Nakagawa, Selichi
SPEECH COMMUNICATION, 2007, 49 (06) : 501 - 513
[42] Enhancing GMM speaker identification by incorporating SVM speaker verification for intelligent web-based speech applications
Ing-Jr Ding
Chih-Ta Yen
Multimedia Tools and Applications, 2015, 74 : 5131 - 5140
[43] A hybrid system based on GMM-SVM for Speaker Identification
Chakroun, Rania
Zouari, Leila Beltaifa
Frikha, Mondher
Ben Hamida, Ahmed
2015 15TH INTERNATIONAL CONFERENCE ON INTELLIGENT SYSTEMS DESIGN AND APPLICATIONS (ISDA), 2015, : 654 - 658
[44] Enhancing GMM speaker identification by incorporating SVM speaker verification for intelligent web-based speech applications
Ding, Ing-Jr
Yen, Chih-Ta
MULTIMEDIA TOOLS AND APPLICATIONS, 2015, 74 (14) : 5131 - 5140
[45] Evaluation of I-vector and GMM Based Speaker Verification Systems for Forensic Application
Gumus, Fatma
Yankayis, Mustafa
Karabiber, Fethullah
2016 24TH SIGNAL PROCESSING AND COMMUNICATION APPLICATION CONFERENCE (SIU), 2016, : 617 - 620
[46] HIERARCHICAL MIXTURE CLUSTERING AND ITS APPLICATION TO GMM BASED TEXT INDEPENDENT SPEAKER IDENTIFICATION
Saeidi, R.
Mohammadi, H. R. Sadegh
Ganchev, T.
Rodman, R. D.
2008 INTERNATIONAL SYMPOSIUM ON TELECOMMUNICATIONS, VOLS 1 AND 2, 2008, : 770 - +
[47] SIGNIFICANCE OF UTTERANCE PARTITIONING IN GMM-SVM BASED SPEAKER VERIFICATION IN VARYING BACKGROUND ENVIRONMENT
Sarkar, Sourjya
Rao, K. Sreenivasa
2013 INTERNATIONAL CONFERENCE ORIENTAL COCOSDA HELD JOINTLY WITH 2013 CONFERENCE ON ASIAN SPOKEN LANGUAGE RESEARCH AND EVALUATION (O-COCOSDA/CASLRE), 2013,
[48] Speech Features Evaluation for Small Set Automatic Speaker Verification Using GMM-UBM System
Rakhmanenko, Ivan
Meshcheryakov, Roman
SPEECH AND COMPUTER, 2016, 9811 : 645 - 650
[49] FUZZY LOGIC-BASED INTELLIGENT CONTROL FOR SVM SPEAKER VERIFICATION WITH THE SUPPORT OF GMM PRIOR INFORMATION
Ding, Ing-Jr
Yen, Chih-Ta
Lin, Zih-Jheng
TRANSACTIONS OF THE CANADIAN SOCIETY FOR MECHANICAL ENGINEERING, 2013, 37 (03) : 467 - 476
[50] Sentence-HMM state-based i-vector/PLDA modelling for improved performance in text dependent single utterance speaker verification
Buyuk, Osman
IET SIGNAL PROCESSING, 2016, 10 (08) : 918 - 923

← 1 2 3 4 5 →