• DocumentCode
    2267808
  • Title

    Articulatory parameter generation using unsupervised Hidden Markov Models

  • Author

    Lachambre, Helene ; Koenig, Lionel ; Andre-Obrecht, Regine

  • Author_Institution
    IRIT, Univ. de Toulouse, Narbonne, France
  • fYear
    2011
  • fDate
    Aug. 29 2011-Sept. 2 2011
  • Firstpage
    456
  • Lastpage
    460
  • Abstract
    We present an acoustic-to-articulatory inversion method based on unsupervised Hidden Markov Models. A global HMM is first trained from the acoustic and articulatory data. This model is then split in two sub-models which represent the acoustic part and the articulatory part of the data. These two sub-models are linked through the fact that they are deduced from the same global model.
  • Keywords
    acoustic signal processing; decoding; estimation theory; hidden Markov models; mean square error methods; speech processing; HMM; RMS error; acoustic vector sequence decoding; acoustic-to-articulatory inversion method; articulatory parameter generation; unsupervised hidden Markov model; Acoustics; Art; Hidden Markov models; Speech; Training; Trajectory; Vectors;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Signal Processing Conference, 2011 19th European
  • Conference_Location
    Barcelona
  • ISSN
    2076-1465
  • Type

    conf

  • Filename
    7074035
    • Link To Document
    https://search.isc.ac/dl/search/defaultta.aspx?DTC=49&DC=2267808