Multi-Task Learning with High-Order Statistics for x-Vector Based Text-Independent Speaker Verification

Lanhua You, Wu Guo, Li-Rong Dai, Jun Du


The x-vector based deep neural network (DNN) embedding systems have demonstrated effectiveness for text-independent speaker verification. This paper presents a multi-task learning architecture for training the speaker embedding DNN with the primary task of classifying the target speakers, and the auxiliary task of reconstructing the first- and higher-order statistics of the original input utterance. The proposed training strategy aggregates both the supervised and unsupervised learning into one framework to make the speaker embeddings more discriminative and robust. Experiments are carried out using the NIST SRE16 evaluation dataset and the VOiCES dataset. The results demonstrate that our proposed method outperforms the original x-vector approach with very low additional complexity added.


 DOI: 10.21437/Interspeech.2019-2264

Cite as: You, L., Guo, W., Dai, L., Du, J. (2019) Multi-Task Learning with High-Order Statistics for x-Vector Based Text-Independent Speaker Verification. Proc. Interspeech 2019, 1158-1162, DOI: 10.21437/Interspeech.2019-2264.


@inproceedings{You2019,
  author={Lanhua You and Wu Guo and Li-Rong Dai and Jun Du},
  title={{Multi-Task Learning with High-Order Statistics for x-Vector Based Text-Independent Speaker Verification}},
  year=2019,
  booktitle={Proc. Interspeech 2019},
  pages={1158--1162},
  doi={10.21437/Interspeech.2019-2264},
  url={http://dx.doi.org/10.21437/Interspeech.2019-2264}
}