Speaker Prediction based on Head Orientations

R.J. Rienks, Ronald Walter Poppe

    Research output: Chapter in Book/Report/Conference proceedingConference contributionAcademicpeer-review

    14 Downloads (Pure)

    Abstract

    To gain insight into gaze behavior in meetings, this paper compares the results from a Naive Bayes classifier, Neural Networks and humans on speaker prediction in four-person meetings given solely the azimuth head angles. The Naive Bayes classifier scored 69.4% correctly, Neural Networks 62.3% and humans only 37.7%. None of the classifiers was able to generalize over meetings. We show that there are strong indications that human specific gaze behavior influences the fact that the models do not generalize. Additionally, we show that for all classifiers the performance of the prediction in the beginning and at the end of a speaker turn is worse than halfway through the speaker turn.
    Original languageUndefined
    Title of host publicationProceedings of the Fourteenth Annual Machine Learning Conference of Belgium and the Netherlands (Benelearn 2005)
    EditorsM. van Otterlo, Mannes Poel, M. Poel, A. Nijholt, Antinus Nijholt
    Place of PublicationEnschede
    PublisherUniversity of Twente
    Pages73-79
    Number of pages7
    Publication statusPublished - 2005
    Event14th Belgium-Dutch Conference on Machine Learning, Benelearn 2005 - Enschede, the Netherlands, Bruges, Belgium
    Duration: 27 Apr 200529 Apr 2005
    Conference number: 14

    Publication series

    NameCTIT Workshop Proceedings Series
    PublisherUniversity of Twente
    VolumeWP05-03
    ISSN (Print)1574-0846

    Conference

    Conference14th Belgium-Dutch Conference on Machine Learning, Benelearn 2005
    Abbreviated titleBenelearn 2005
    CountryBelgium
    CityBruges
    Period27/04/0529/04/05
    OtherFebruary 17-18, 2005

    Keywords

    • EWI-1815
    • METIS-227355
    • IR-65556
    • EC Grant Agreement nr.: FP6/506811

    Cite this