Corpus ID: 38080504

Improved topic classification and keyword discovery using an HMM-based speech recognizer trained without supervision

@inproceedings{Siu2010ImprovedTC,
  title={Improved topic classification and keyword discovery using an HMM-based speech recognizer trained without supervision},
  author={M. Siu and H. Gish and A. Chan and W. Belfield},
  booktitle={INTERSPEECH},
  year={2010}
}
  • M. Siu, H. Gish, +1 author W. Belfield
  • Published in INTERSPEECH 2010
  • Computer Science
  • In our previous publication [1], we presented a new approach to HMM training, viz., training without supervision. We used an HMM trained without supervision for transcribing audio into self-organized units (SOUs) for the purpose of topic classification. In this paper we report improvements made to the system, including the use of context dependent acoustic models and lattice based features that together reduce the topic verification equal error rate from 12% to 7%. In addition to discussing the… CONTINUE READING
    Efficient spoken term discovery using randomized algorithms
    • 130
    • PDF
    Towards Unsupervised Training of Speaker Independent Acoustic Models
    • 73
    • PDF
    Towards unsupervised speech processing
    • 42
    • PDF

    References

    Publications referenced by this paper.
    SHOWING 1-10 OF 13 REFERENCES
    LIBSVM: A library for support vector machines
    • 36,428
    • PDF
    A segmental speech model with applications to word spotting
    • 128
    Subspace Gaussian Mixture Models for speech recognition
    • 174
    • PDF
    Topic identification from audio recordings using word and phone recognition lattices
    • 46
    • PDF
    Towards multi-speaker unsupervised speech pattern discovery
    • 78
    • PDF
    Discriminative Keyword Selection Using Support Vector Machines
    • 16
    • PDF
    Parametric trajectory models for speech recognition
    • 101
    • PDF
    Advances in transcription of broadcast news and conversational telephone speech within the combined EARS BBN/LIMSI system
    • 54
    • Highly Influential
    • PDF
    Techniques for rapid and robust topic identification of conversational telephone speech
    • 15
    • PDF