Four weightings and a fusion: a cepstral-SVM system for speaker recognition

@article{Kajarekar2005FourWA,
  title={Four weightings and a fusion: a cepstral-SVM system for speaker recognition},
  author={S. Kajarekar},
  journal={IEEE Workshop on Automatic Speech Recognition and Understanding, 2005.},
  year={2005},
  pages={17-22}
}
  • S. Kajarekar
  • Published 2005
  • Computer Science
  • IEEE Workshop on Automatic Speech Recognition and Understanding, 2005.
  • A new speaker recognition system is described that uses Mel-frequency cepstral features. This system is a combination of four support vector machines (SVMs). All the SVM systems use polynomial features and they are trained and tested independently using a linear inner-product kernel. Scores from each system are combined with equal weight to generate the final score. We evaluate the combined SVM system using extensive development sets with diverse recording conditions. These sets include NIST… CONTINUE READING

    Figures, Tables, and Topics from this paper.

    Phone-based Cepstral Polynomial SVM System for Speaker Recognition
    • 10
    • Highly Influenced
    • PDF
    Phone-based cepstral polynomial SVM system for speaker recognition
    AN SVM FRONT-END LANDMARK SPEECH RECOGNITION SYSTEM
    • 6
    • PDF
    Speaker Recognition With Session Variability Normalization Based on MLLR Adaptation Transforms
    • 80
    • Highly Influenced
    • PDF
    Kernel combination for SVM speaker verification
    • 17
    • PDF
    Improvements in MLLR-Transform-based Speaker Recognition
    • 36
    • PDF
    The Contribution of Cepstral and Stylistic Features to SRI's 2005 NIST Speaker Recognition Evaluation System
    • 25
    • PDF
    Within-class covariance normalization for SVM-based speaker recognition
    • 426
    • Highly Influenced
    • PDF
    Higher-Level Features in Speaker Recognition
    • 80
    • PDF

    References

    Publications referenced by this paper.
    SHOWING 1-10 OF 12 REFERENCES
    MLLR transforms as features in speaker recognition
    • 157
    • PDF
    High-level speaker verification with support vector machines
    • 83
    • PDF
    Generalized linear discriminant sequence kernels for speaker recognition
    • 230
    • Highly Influential
    Improving a GMM speaker verification system by phonetic weighting
    • 63
    • Highly Influential
    Speaker Verification Using Adapted Gaussian Mixture Models
    • 4,378
    • Highly Influential
    • PDF
    SVM modeling of "SNERF-grams" for speaker recognition
    • 19
    • PDF
    Channel compensation for SVM speaker recognition
    • 155
    • PDF
    Improved phonetic speaker recognition using lattice decoding
    • 59
    • PDF
    Speaker identification and verification using eigenvoices
    • 60
    • PDF