Non-linear PLDA for i-Vector Speaker Verification

被引：0
|
作者：
Novoselov, Sergey [1 ]
Pekhovsky, Timur [1 ,2 ]
Kudashev, Oleg [1 ]
Mendelev, Valentin [1 ,2 ]
Prudnikov, Alexey [1 ]
机构：
[1] Speech Technol Ctr Ltd, St Petersburg, Russia
[2] ITMO Univ, St Petersburg, Russia
来源：
16TH ANNUAL CONFERENCE OF THE INTERNATIONAL SPEECH COMMUNICATION ASSOCIATION (INTERSPEECH 2015), VOLS 1-5 | 2015年
关键词：
i-vector; PLDA; RBM; autoencoder; DDML;
D O I：
暂无
中图分类号：
O42 [声学];
学科分类号：
070206 ; 082403 ;
摘要：
Two approaches are presented for non-linear PLDA to be used in speaker verification. In NIST 2010 speaker recognition evaluation (SRE) tests under DET-5 conditions, the two methods and particularly their combination provided significant improvements in equal error rates and minDCF values over a standard PLDA scheme. The proposed schemes were also applied within a speaker verification system that employs DNN-based sufficient statistics calculation resulting in a 45 % reduction in minDCF relative to a conventional GMM based system
引用
收藏
页码：214 / 218
页数：5
相关论文
共 50 条
  • [1] Fast Scoring for Mixture of PLDA in I-Vector/PLDA Speaker Verification
    Mak, Man-Wai
    [J]. 2015 ASIA-PACIFIC SIGNAL AND INFORMATION PROCESSING ASSOCIATION ANNUAL SUMMIT AND CONFERENCE (APSIPA), 2015, : 587 - 593
  • [2] PLDA Modeling in I-Vector and Supervector Space for Speaker Verification
    Jiang, Ye
    Lee, Kong Aik
    Tang, Zhenmin
    Ma, Bin
    Larcher, Anthony
    Li, Haizhou
    [J]. 13TH ANNUAL CONFERENCE OF THE INTERNATIONAL SPEECH COMMUNICATION ASSOCIATION 2012 (INTERSPEECH 2012), VOLS 1-3, 2012, : 1678 - 1681
  • [3] Nonparametrically trained PLDA for short duration i-vector speaker verification
    Khosravani, Abbas
    Homayounpour, Mohammad M.
    [J]. COMPUTER SPEECH AND LANGUAGE, 2018, 52 : 105 - 122
  • [4] NORMALIZATION OF TOTAL VARIABILITY MATRIX FOR I-VECTOR/PLDA SPEAKER VERIFICATION
    Rao, Wei
    Mak, Man-Wai
    Lee, Kong-Aik
    [J]. 2015 IEEE INTERNATIONAL CONFERENCE ON ACOUSTICS, SPEECH, AND SIGNAL PROCESSING (ICASSP), 2015, : 4180 - 4184
  • [5] I-VECTOR KULLBACK-LEIBLER DIVISIVE NORMALIZATION FOR PLDA SPEAKER VERIFICATION
    Pan, Yilin
    Zheng, Tieran
    Chen, Chen
    [J]. 2017 IEEE GLOBAL CONFERENCE ON SIGNAL AND INFORMATION PROCESSING (GLOBALSIP 2017), 2017, : 56 - 60
  • [6] DEEP NEURAL NETWORK DRIVEN MIXTURE OF PLDA FOR ROBUST I-VECTOR SPEAKER VERIFICATION
    Li, Na
    Mak, Man-Wai
    Chien, Jen-Tzung
    [J]. 2016 IEEE WORKSHOP ON SPOKEN LANGUAGE TECHNOLOGY (SLT 2016), 2016, : 186 - 191
  • [7] DEEP NEURAL NETWORK BASED DISCRIMINATIVE TRAINING FOR I-VECTOR/PLDA SPEAKER VERIFICATION
    Zheng Tieran
    Han Jiqing
    Zheng Guibin
    [J]. 2018 IEEE INTERNATIONAL CONFERENCE ON ACOUSTICS, SPEECH AND SIGNAL PROCESSING (ICASSP), 2018, : 5354 - 5358
  • [8] Turkish Text-Dependent Speaker Verification using i-vector/PLDA Approach
    Hanilci, Cemal
    Celiktas, Havva
    [J]. 2018 26TH SIGNAL PROCESSING AND COMMUNICATIONS APPLICATIONS CONFERENCE (SIU), 2018,
  • [9] NON-PARALLEL VOICE CONVERSION USING I-VECTOR PLDA: TOWARDS UNIFYING SPEAKER VERIFICATION AND TRANSFORMATION
    Kinnunen, Tomi
    Juvela, Lauri
    Alku, Paavo
    Yamagishi, Junichi
    [J]. 2017 IEEE INTERNATIONAL CONFERENCE ON ACOUSTICS, SPEECH AND SIGNAL PROCESSING (ICASSP), 2017, : 5535 - 5539
  • [10] FULL-COVARIANCE UBM AND HEAVY-TAILED PLDA IN I-VECTOR SPEAKER VERIFICATION
    Matejka, Pavel
    Glembek, Ondrej
    Castaldo, Fabio
    Alam, M. J.
    Plchot, Oldrich
    Kenny, Patrick
    Burget, Lukas
    Cernocky, Jan 'Honza'
    [J]. 2011 IEEE INTERNATIONAL CONFERENCE ON ACOUSTICS, SPEECH, AND SIGNAL PROCESSING, 2011, : 4828 - 4831
12345