We present a novel approach for full body pose tracking using stochastic sampling. A volumetric reconstruction of a person is extracted from silhouettes in multiple video images. Then, an articulated body model is fitted to the data with stochastic meta descent (SMD) optimization. By comparing even a simplified version of SMD to the commonly used Levenberg-Marquardt method, we demonstrate the power of stochastic compared to deterministic sampling, especially in cases of noisy and incomplete data. Moreover, color information is added to improve the speed and robustness of the tracking. Results are shown for several challenging sequences, with tracking of 24 degrees of freedom in less than 1 second per frame.


    Access

    Check access

    Check availability in my library

    Order at Subito €


    Export, share and cite



    Title :

    Full body tracking from multiple views using stochastic sampling


    Contributors:
    Kehl, R. (author) / Bray, M. (author) / Van Gool, L. (author)


    Publication date :

    2005-01-01


    Size :

    843813 byte





    Type of media :

    Conference paper


    Type of material :

    Electronic Resource


    Language :

    English



    Markerless tracking of complex human motions from multiple views

    Kehl, R. / Gool, L. V. | British Library Online Contents | 2006


    Tracking across multiple cameras with disjoint views

    Javed, / Rasheed, / Shafique, et al. | IEEE | 2003


    3D Human Body Model Acquisition from Multiple Views

    Kakadiaris, I. / Metaxas, D. / IEEE Computer Society et al. | British Library Conference Proceedings | 1995


    Tracking across Multiple Cameras with Disjoint Views

    Javed, O. / Rasheed, Z. / Shafique, K. et al. | British Library Conference Proceedings | 2003


    Tracking Hybrid 2D-3D Human Models from Multiple Views

    Ong, E. / Gong, S. / IEEE Computer Society | British Library Conference Proceedings | 1999