ISCA Archive Eurospeech 1997
ISCA Archive Eurospeech 1997

GMM sample statistic log-likelihoods for text-independent speaker recognition

Michael Schmidt, John Golden, Herbert Gish

A novel approach to scoring Gaussian mixture mod- els is presented. Feature vectors are assigned to the individual Gaussians making up the model and log-likelihoods of the separate Gaussians are computed and summed. Furthermore, the log-likelihoods of the individual Gaussians can be decomposed into sample weight, mean, and covariance log-likelihoods. Correlation likelihoods can also be computed. The results of the various systems are comparable on text- independent speaker recognition experiments despite the fact that the models and scoring are all quite di erent. By decomposing log-likelihoods of models into various sample statistic log-likelihoods, it is possible to diagnose which part of the model has the greatest discriminative power, whether the location of the Gaussians or their shapes.


doi: 10.21437/Eurospeech.1997-288

Cite as: Schmidt, M., Golden, J., Gish, H. (1997) GMM sample statistic log-likelihoods for text-independent speaker recognition. Proc. 5th European Conference on Speech Communication and Technology (Eurospeech 1997), 855-858, doi: 10.21437/Eurospeech.1997-288

@inproceedings{schmidt97_eurospeech,
  author={Michael Schmidt and John Golden and Herbert Gish},
  title={{GMM sample statistic log-likelihoods for text-independent speaker recognition}},
  year=1997,
  booktitle={Proc. 5th European Conference on Speech Communication and Technology (Eurospeech 1997)},
  pages={855--858},
  doi={10.21437/Eurospeech.1997-288},
  issn={1018-4074}
}