Fuzzy Modelling Techniques for Speech Recognition

Dat Tran, Max Wagner

    Research output: A Conference proceeding or a Chapter in BookConference contribution

    Abstract

    A fuzzy modelling approach to hidden Markov models (HMMs) for speech recognition is presented in this paper. State sequences are viewed as fuzzy sets and a timedependent fuzzy membership function is defined to represent the degree of belonging of an observation sequence to fuzzy state sequences at each time. An optimisation criterion for fuzzy HMMs is also presented. Parameter reestimation equations are derived from this criterion and used to train fuzzy HMMs for speech recognition. Fuzzy models are more effective than conventional models in isolated word recognition performed on the TI46 speech data corpus.
    Original languageEnglish
    Title of host publicationProceedings of the Ninth Australian International Conference on Speech Science and Technology
    Place of PublicationMelbourne, Australia
    PublisherAustralian Speech Science and Technology Associatn
    Pages473-478
    Number of pages6
    ISBN (Print)0 9581946 0 2
    Publication statusPublished - 2002
    EventNinth Australian International Conference on Speech Science and Technology - Melbourne, Australia
    Duration: 3 Dec 20025 Dec 2002

    Conference

    ConferenceNinth Australian International Conference on Speech Science and Technology
    CountryAustralia
    CityMelbourne
    Period3/12/025/12/02

    Fingerprint

    Hidden Markov models
    Speech recognition
    Membership functions
    Fuzzy sets

    Cite this

    Tran, D., & Wagner, M. (2002). Fuzzy Modelling Techniques for Speech Recognition. In Proceedings of the Ninth Australian International Conference on Speech Science and Technology (pp. 473-478). Melbourne, Australia: Australian Speech Science and Technology Associatn.
    Tran, Dat ; Wagner, Max. / Fuzzy Modelling Techniques for Speech Recognition. Proceedings of the Ninth Australian International Conference on Speech Science and Technology. Melbourne, Australia : Australian Speech Science and Technology Associatn, 2002. pp. 473-478
    @inproceedings{3d3f9ee48c6040d3bd2c21a539a10afe,
    title = "Fuzzy Modelling Techniques for Speech Recognition",
    abstract = "A fuzzy modelling approach to hidden Markov models (HMMs) for speech recognition is presented in this paper. State sequences are viewed as fuzzy sets and a timedependent fuzzy membership function is defined to represent the degree of belonging of an observation sequence to fuzzy state sequences at each time. An optimisation criterion for fuzzy HMMs is also presented. Parameter reestimation equations are derived from this criterion and used to train fuzzy HMMs for speech recognition. Fuzzy models are more effective than conventional models in isolated word recognition performed on the TI46 speech data corpus.",
    author = "Dat Tran and Max Wagner",
    year = "2002",
    language = "English",
    isbn = "0 9581946 0 2",
    pages = "473--478",
    booktitle = "Proceedings of the Ninth Australian International Conference on Speech Science and Technology",
    publisher = "Australian Speech Science and Technology Associatn",

    }

    Tran, D & Wagner, M 2002, Fuzzy Modelling Techniques for Speech Recognition. in Proceedings of the Ninth Australian International Conference on Speech Science and Technology. Australian Speech Science and Technology Associatn, Melbourne, Australia, pp. 473-478, Ninth Australian International Conference on Speech Science and Technology, Melbourne, Australia, 3/12/02.

    Fuzzy Modelling Techniques for Speech Recognition. / Tran, Dat; Wagner, Max.

    Proceedings of the Ninth Australian International Conference on Speech Science and Technology. Melbourne, Australia : Australian Speech Science and Technology Associatn, 2002. p. 473-478.

    Research output: A Conference proceeding or a Chapter in BookConference contribution

    TY - GEN

    T1 - Fuzzy Modelling Techniques for Speech Recognition

    AU - Tran, Dat

    AU - Wagner, Max

    PY - 2002

    Y1 - 2002

    N2 - A fuzzy modelling approach to hidden Markov models (HMMs) for speech recognition is presented in this paper. State sequences are viewed as fuzzy sets and a timedependent fuzzy membership function is defined to represent the degree of belonging of an observation sequence to fuzzy state sequences at each time. An optimisation criterion for fuzzy HMMs is also presented. Parameter reestimation equations are derived from this criterion and used to train fuzzy HMMs for speech recognition. Fuzzy models are more effective than conventional models in isolated word recognition performed on the TI46 speech data corpus.

    AB - A fuzzy modelling approach to hidden Markov models (HMMs) for speech recognition is presented in this paper. State sequences are viewed as fuzzy sets and a timedependent fuzzy membership function is defined to represent the degree of belonging of an observation sequence to fuzzy state sequences at each time. An optimisation criterion for fuzzy HMMs is also presented. Parameter reestimation equations are derived from this criterion and used to train fuzzy HMMs for speech recognition. Fuzzy models are more effective than conventional models in isolated word recognition performed on the TI46 speech data corpus.

    M3 - Conference contribution

    SN - 0 9581946 0 2

    SP - 473

    EP - 478

    BT - Proceedings of the Ninth Australian International Conference on Speech Science and Technology

    PB - Australian Speech Science and Technology Associatn

    CY - Melbourne, Australia

    ER -

    Tran D, Wagner M. Fuzzy Modelling Techniques for Speech Recognition. In Proceedings of the Ninth Australian International Conference on Speech Science and Technology. Melbourne, Australia: Australian Speech Science and Technology Associatn. 2002. p. 473-478