Task-Driven Variability Model for Speaker Verification
Task-Driven Variability Model for Speaker Verification
复制标题
用于说话人验证的任务驱动变异模型
DOI：
10.1007/s00034-019-01315-7
复制
发表时间：
2019-11
期刊：
影响因子：
--
通讯作者：
Han Jiqing
中科院分区：
文献类型：
--
作者：
Chen Chen;Han Jiqing
The total variability model (TVM)/probabilistic linear discriminant analysis (PLDA) framework is one of the most popular methods for speaker verification. In this framework, the i-vector representations are first extracted from utterances via an estimated TVM and then employed to estimate the PLDA parameters for classification. The TVM and PLDA are estimated serially, so the information loss in the TVM is inherited by the i-vectors, and then passed into the PLDA classifier. More seriously, this loss cannot be compensated by the PLDA. To solve this problem, we propose a task-driven variability model (TDVM) to jointly estimate the TVM and PLDA classifier. In this method, the feedback from the PLDA can supervise the optimal solution of the TVM to move toward the space that has the maximum between-class separation and minimum within-class variation. Meanwhile, this space is suitable for open-set test which can deal with unenrolled speakers. Unlike most embedding methods which extract the embedding representations via the stack of network structures, the TDVM contains the assumptions about latent variables, which can enhance the interpretation of speaker representation extraction. The proposed method is evaluated on the King-ASR-010 and VoxCeleb databases, and the experimental results show that the TDVM method can achieve better performance than the traditional TVM/PLDA and VGG-M network with different cost functions.
登录
查看更多内容
DOI：
10.1007/s00034-015-0206-2
发表时间：
2015-12
期刊：
Circuits, Systems, and Signal Processing
影响因子：
--
作者：
Yan Xu;I. Mcloughlin;Yan Song;Kui Wu
通讯作者：
Yan Xu;I. Mcloughlin;Yan Song;Kui Wu
DOI：
10.21437/interspeech.2017-633
发表时间：
2017-08
期刊：
--
影响因子：
--
作者：
Chen-Chen-Chen;Jiqing Han;Yilin Pan
通讯作者：
Chen-Chen-Chen;Jiqing Han;Yilin Pan
DOI：
10.1109/taslp.2016.2515506
发表时间：
2016-03
期刊：
IEEE/ACM Transactions on Audio, Speech, and Language Processing
影响因子：
--
作者：
S. Shepstone;Kong Aik LEE;Haizhou Li;Z. Tan;S. H. Jensen
通讯作者：
S. Shepstone;Kong Aik LEE;Haizhou Li;Z. Tan;S. H. Jensen
DOI：
10.1109/taslp.2018.2806305
发表时间：
2018-05
期刊：
IEEE/ACM Transactions on Audio, Speech, and Language Processing
影响因子：
--
作者：
Sandro Cumani;P. Laface
通讯作者：
Sandro Cumani;P. Laface
DOI：
10.1109/eusipco.2015.7362754
发表时间：
2015-12
期刊：
2015 23rd European Signal Processing Conference (EUSIPCO)
影响因子：
--
作者：
Y. Isik;Hakan Erdogan;R. Sarikaya
通讯作者：
Y. Isik;Hakan Erdogan;R. Sarikaya