This paper describes a fully unsupervised approach to speaker clustering and labeling employing speech recognition (ASR) technology to bootstrap speaker identification (SID). An algorithm that combined these two technologies was able to correctly cluster and label 299 NATO ship-to-ship transmissions with an accuracy of 89% in an on-line (no a priori training) scenario. This fusion approach out-performed ASR alone by 23.6%, and outperformed manually-trained VQ-SID by 12.7% and GMM/UMB-SID by 8.6%. This paper demonstrates that, under certain circumstances, unsupervised, self-organizing systems can be more effective than manually-trained ones.


    Access

    Check access

    Check availability in my library

    Order at Subito €


    Export, share and cite



    Title :

    Automatic speech recognition fusion approach to unsupervised speaker clustering and labeling


    Contributors:
    Lawson, A.D. (author) / Huggins, M.C. (author) / Grieco, J.J. (author) / Galligan, S.A. (author) / Harris, D.M. (author)

    Published in:

    Publication date :

    2006-01-01


    Size :

    188867 byte




    Type of media :

    Conference paper


    Type of material :

    Electronic Resource


    Language :

    English




    Dignet: an unsupervised-learning clustering algorithm for clustering and data fusion

    Thomopoulos, S.C.A. / Bougoulias, D.K. / Chin-Der Wann, | IEEE | 1995


    Speech variability in automatic speaker recognition systems for commercial and forensic purposes

    Ortega-Garcia, J. / Gonzalez-Rodriguez, J. / Cruz-Llanas, S. | IEEE | 2000


    Mathematical Analysis and Speaker-Independent Speech Recognition

    Gur'yanov, A. E. | British Library Online Contents | 1996


    Articles - Automatic Speaker Recognition Variability

    Ortega-Garcia | Online Contents | 2000