Large vocabulary conversational speech recognition with a subspace constraint on inverse covariance matrices

@inproceedings{Axelrod2003LargeVC,
  title={Large vocabulary conversational speech recognition with a subspace constraint on inverse covariance matrices},
  author={Scott Axelrod and Vaibhava Goel and Brian Kingsbury and Karthik Visweswariah and Ramesh A. Gopinath},
  booktitle={INTERSPEECH},
  year={2003}
}
This paper applies the recently proposed SPAM models for acoustic modeling in a Speaker Adaptive Training (SAT) context on large vocabulary conversational speech databases, including the Switchboard database. SPAM models are Gaussian mixture models in which a subspace constraint is placed on the precision and mean matrices (although this paper focuses on the case of unconstrained means). They include diagonal covariance, full covariance, MLLT, and EMLLT models as special cases. Adaptation is… CONTINUE READING

Citations

Publications citing this paper.
Showing 1-9 of 9 extracted citations

SPAM and full covariance for speech recognition

INTERSPEECH • 2006
View 5 Excerpts
Highly Influenced

Adaptation of precision matrix models on large vocabulary continuous speech recognition

Proceedings. (ICASSP '05). IEEE International Conference on Acoustics, Speech, and Signal Processing, 2005. • 2005
View 6 Excerpts
Highly Influenced

Investigation of acoustic modeling techniques for LVCSR systems

Proceedings. (ICASSP '05). IEEE International Conference on Acoustics, Speech, and Signal Processing, 2005. • 2005
View 5 Excerpts
Highly Influenced

State of the art discriminative training of subspace constrained Gaussian mixture models in big training corpora

2013 IEEE International Conference on Acoustics, Speech and Signal Processing • 2013
View 1 Excerpt

Minimum phone error training of precision matrix models

IEEE Transactions on Audio, Speech, and Language Processing • 2006
View 2 Excerpts

References

Publications referenced by this paper.
Showing 1-10 of 17 references

Semi-tied covariance matrices for hidden Markov models

IEEE Trans. Speech and Audio Processing • 1999
View 9 Excerpts
Highly Influenced

Modeling inverse covariance matrices by basis expansion

IEEE Transactions on Speech and Audio Processing • 2002

Similar Papers

Loading similar papers…