ISCA Archive Interspeech 2013
ISCA Archive Interspeech 2013

Automatic regularization of cross-entropy cost for speaker recognition fusion

Ville Hautamäki, Kong Aik Lee, David A. van Leeuwen, R. Saeidi, Anthony Larcher, Tomi Kinnunen, Taufiq Hasan, Seyed Omid Sadjadi, Gang Liu, Hynek Bořil, John H. L. Hansen, Benoit Fauve

In this paper we study automatic regularization techniques for the fusion of automatic speaker recognition systems. Parameter regularization could dramatically reduce the fusion training time. In addition, there will not be any need for splitting the development set into different folds for cross- validation. We utilize majorization-minimization approach to automatic ridge regression learning and design a similar way to learn LASSO regularization parameter automatically. By experiments we show improvement in using automatic regularization.

doi: 10.21437/Interspeech.2013-404

Cite as: Hautamäki, V., Lee, K.A., Leeuwen, D.A.v., Saeidi, R., Larcher, A., Kinnunen, T., Hasan, T., Sadjadi, S.O., Liu, G., Bořil, H., Hansen, J.H.L., Fauve, B. (2013) Automatic regularization of cross-entropy cost for speaker recognition fusion. Proc. Interspeech 2013, 1609-1613, doi: 10.21437/Interspeech.2013-404

  author={Ville Hautamäki and Kong Aik Lee and David A. van Leeuwen and R. Saeidi and Anthony Larcher and Tomi Kinnunen and Taufiq Hasan and Seyed Omid Sadjadi and Gang Liu and Hynek Bořil and John H. L. Hansen and Benoit Fauve},
  title={{Automatic regularization of cross-entropy cost for speaker recognition fusion}},
  booktitle={Proc. Interspeech 2013},