An approach is proposed for clustering time-series data. The approach can be used to discover groupings of similar object motions that were observed in a video collection. A finite mixture of hidden Markov models (HMMs) is fitted to the motion data using the expectation maximization (EM) framework. Previous approaches for HMM-based clustering employ a k-means formulation, where each sequence is assigned to only a single HMM. In contrast, the formulation presented in this paper allows each sequence to belong to more than a single HMM with some probability, and the hard decision about the sequence class membership can be deferred until a later time when such a decision is required. Experiments with simulated data demonstrate the benefit of using this EM-based approach when there is more "overlap" in the processes generating the data. Experiments with real data show the promising potential of HMM-based motion clustering in a number of applications.


    Zugriff

    Zugriff prüfen

    Verfügbarkeit in meiner Bibliothek prüfen

    Bestellung bei Subito €


    Exportieren, teilen und zitieren



    Titel :

    Discovering clusters in motion time-series data


    Beteiligte:
    Alon, J. (Autor:in) / Sclaroff, S. (Autor:in) / Kollios, G. (Autor:in) / Pavlovic, V. (Autor:in)


    Erscheinungsdatum :

    01.01.2003


    Format / Umfang :

    371521 byte





    Medientyp :

    Aufsatz (Konferenz)


    Format :

    Elektronische Ressource


    Sprache :

    Englisch



    Discovering Clusters in Motion Time-Series Data

    Alon, J. / Sclaroff, S. / Kollios, G. et al. | British Library Conference Proceedings | 2003


    Discovering instantaneous and lagged causal relations in autocorrelated nonlinear time series datasets

    Runge, Jakob | Deutsches Zentrum für Luft- und Raumfahrt (DLR) | 2020

    Freier Zugriff

    Discovering Engineering

    McMurtry, Kate | NTRS | 2015


    Discovering Mars

    NTRS | 1992


    Discovering Safety Issues by Combining Data Sources

    Records, R. M. / Dobias, O. C. / Marsteller, J. M. et al. | British Library Conference Proceedings | 1998