Corpus ID: 6820627

Full Covariance Modelling for Speech Recognition

@inproceedings{Bell2010FullCM,
  title={Full Covariance Modelling for Speech Recognition},
  author={Peter Bell},
  year={2010}
}
  • Peter Bell
  • Published 2010
  • Computer Science
  • HMM-based systems for Automatic Speech Recognition typically model the acoustic features using mixtures of multivariate Gaussians. In this thesis, we consider the problem of learning a suitable covariance matrix for each Gaussian. A variety of schemes have been proposed for controlling the number of covariance parameters per Gaussian, and studies have shown that in general, the greater the number of parameters used in the models, the better the recognition performance. We therefore investigate… CONTINUE READING

    Citations

    Publications citing this paper.
    SHOWING 1-10 OF 13 CITATIONS

    Sparse Inverse Covariance Matrices for Low Resource Speech Recognition

    VIEW 1 EXCERPT
    CITES METHODS

    Discriminatively Trained Sparse Inverse Covariance Matrices for Speech Recognition

    Subspace models for bottleneck features

    VIEW 2 EXCERPTS
    CITES BACKGROUND

    Efficient Sparse Banded Acoustic Models for Speech Recognition

    VIEW 1 EXCERPT
    CITES BACKGROUND

    Tying rotations of covariance matrices via riemannian subspace clustering

    • Yusuke Shinohara
    • Mathematics, Computer Science
    • 2013 IEEE International Conference on Acoustics, Speech and Signal Processing
    • 2013
    VIEW 1 EXCERPT
    CITES BACKGROUND

    References

    Publications referenced by this paper.
    SHOWING 1-10 OF 57 REFERENCES

    Factor analysed hidden Markov models

    VIEW 5 EXCERPTS
    HIGHLY INFLUENTIAL

    Large margin hidden Markov models for speech recognition

    Subspace Gaussian Mixture Models for speech recognition

    VIEW 3 EXCERPTS
    HIGHLY INFLUENTIAL

    Structurally discriminative graphical models for automatic speech recognition - results from the 2001 Johns Hopkins Summer Workshop

    A Maximum Likelihood Approach to Continuous Speech Recognition

    VIEW 10 EXCERPTS
    HIGHLY INFLUENTIAL

    Maximum mutual information based reduction strategies for cross-correlation based joint distributional modeling

    • Jeff A. Bilmes
    • Computer Science
    • Proceedings of the 1998 IEEE International Conference on Acoustics, Speech and Signal Processing, ICASSP '98 (Cat. No.98CH36181)
    • 1998
    VIEW 3 EXCERPTS
    HIGHLY INFLUENTIAL