To reuse existing motion data and generate new motion, a method of human motion nonlinear dimensionality reduction and generation, based on fast adaptive scaled Gaussian process latent variable models, is proposed. Through statistical learning on motion data, the motion data are mapped from high-dimensional observation space to low-dimensional latent space to implement nonlinear dimensionality reduction, and probability distributing of posture space which measures the nature of posture is obtained. The posture which meets constraints and has maximal probability can be computed as the solution of inverse kinematics. This method can avoid cockamamie computation and posture distortion existing in traditional inverse kinematics.The experiments show that our method has higher convergence velocity and precision and extends editing range of motion by adapting motion editing direction.


    Access

    Download


    Export, share and cite



    Title :

    Dimensionality Reduction and Generation of Human Motion


    Contributors:
    Qu, Shi (author) / Wu, Ling-da (author) / Yu, Rong-huan (author) / Wei, Ying-mei (author)

    Publication date :

    2013-11-11


    Remarks:

    doi:10.15837/ijccc.2013.6.763
    INTERNATIONAL JOURNAL OF COMPUTERS COMMUNICATIONS & CONTROL; Vol 8 No 6 (2013): International Journal of Computers Communications & Control (December); 869-877 ; 1841-9844 ; 1841-9836 ; 10.15837/ijccc.2013.6



    Type of media :

    Article (Journal)


    Type of material :

    Electronic Resource


    Language :

    English



    Classification :

    DDC:    629






    Dimensionality Reduction for Face Recognition

    Martinez, A. / Vitria, J. / University of Naples et al. | British Library Conference Proceedings | 1997


    An HOG-LGBPHS human detector with dimensionality reduction by PLS

    Bin, H. / Chunxia, Z. | British Library Online Contents | 2014


    Dimensionality Reduction of Sliced-Normal Distributions

    Luis G Crespo / Brendon K Colbert / Sean P Kenny et al. | NTRS