State-of-the-Art Speaker Recognition for Telephone and Video Speech: The JHU-MIT Submission for NIST SRE18

Jesús Villalba, Nanxin Chen, David Snyder, Daniel Garcia-Romero, Alan McCree, Gregory Sell, Jonas Borgstrom, Fred Richardson, Suwon Shon, François Grondin, Réda Dehak, Leibny Paola García-Perera, Daniel Povey, Pedro A. Torres-Carrasquillo, Sanjeev Khudanpur, Najim Dehak


We present a condensed description of the joint effort of JHU-CLSP, JHU-HLTCOE, MIT-LL., MIT CSAIL and LSE-EPITA for NIST SRE18. All the developed systems consisted of x-vector/i-vector embeddings with some flavor of PLDA backend. Very deep x-vector architectures — Extended and Factorized TDNN, and ResNets — clearly outperformed shallower x-vectors and i-vectors. The systems were tailored to the video (VAST) or to the telephone (CMN2) condition. The VAST data was challenging, yielding 4 times worse performance than other video based datasets like Speakers in the Wild. We were able to calibrate the VAST data with very few development trials by using careful adaptation and score normalization methods. The VAST primary fusion yielded EER=10.18% and Cprimary=0.431. By improving calibration in post-eval, we reached Cprimary=0.369. In CMN2, we used unsupervised SPLDA adaptation based on agglomerative clustering and score normalization to correct the domain shift between English and Tunisian Arabic models. The CMN2 primary fusion yielded EER=4.5% and Cprimary=0.313. Extended TDNN x-vector was the best single system obtaining EER=11.1% and Cprimary=0.452 in VAST; and 4.95% and 0.354 in CMN2.


 DOI: 10.21437/Interspeech.2019-2713

Cite as: Villalba, J., Chen, N., Snyder, D., Garcia-Romero, D., McCree, A., Sell, G., Borgstrom, J., Richardson, F., Shon, S., Grondin, F., Dehak, R., García-Perera, L.P., Povey, D., Torres-Carrasquillo, P.A., Khudanpur, S., Dehak, N. (2019) State-of-the-Art Speaker Recognition for Telephone and Video Speech: The JHU-MIT Submission for NIST SRE18. Proc. Interspeech 2019, 1488-1492, DOI: 10.21437/Interspeech.2019-2713.


@inproceedings{Villalba2019,
  author={Jesús Villalba and Nanxin Chen and David Snyder and Daniel Garcia-Romero and Alan McCree and Gregory Sell and Jonas Borgstrom and Fred Richardson and Suwon Shon and François Grondin and Réda Dehak and Leibny Paola García-Perera and Daniel Povey and Pedro A. Torres-Carrasquillo and Sanjeev Khudanpur and Najim Dehak},
  title={{State-of-the-Art Speaker Recognition for Telephone and Video Speech: The JHU-MIT Submission for NIST SRE18}},
  year=2019,
  booktitle={Proc. Interspeech 2019},
  pages={1488--1492},
  doi={10.21437/Interspeech.2019-2713},
  url={http://dx.doi.org/10.21437/Interspeech.2019-2713}
}