Dimensionality Reduction and Generation of Human Motion


  • Shi Qu Air Force Early Warning Academy No.288, Huangpu Road, Wuhan, 430019, China
  • Ling-da Wu School of Equipment, Beijing, 101400, China
  • Rong-huan Yu
  • Ying-mei Wei


dimensionality reduction, Gaussian process, kernel function, motion generation


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.


Sun Hongwei, Gu Ming and Sun Jiaguang,A coding algorithm using PCA-based correlation vector quantization, Journal of Ccomputer-Aided Design & Computer Graphics,17(8):1662- 1666,2005.

Gift N, Lorraine B and Isobel C G, Probabilistic principal component analysis for metabolomic data, BMC Bioinformatics,11(1):571-582,2010. http://dx.doi.org/10.1186/1471-2105-11-571

Nisbet R, Elder J and Miner G, Statistical analysis and data mining, New York:Academic Press,2009.

Roman Rosipal et al, Kernel PCA for feature extraction and de-noising in non-linear regression, Neural Computing & Applications,10(3):231- 243,2001. http://dx.doi.org/10.1007/s521-001-8051-z

Carl H, Philip H S and Neil D L, Gaussian process latent variable models for human pose estimation, Proc. of Machine Learning for Multimodal Interaction. Brno: Springer-Verlag Press:132-143,2007.

QU Shi et al, Pose Synthesis of Virtual Character Based on Statistical Learning, The International Symposium on Computer Network and Multimedia, Wuhan, China:36-39,2009.

Keith Grochow et al, Style-based inverse kinematics, ACM Transactions on Graphics,23(3): 522-531,2004. http://dx.doi.org/10.1145/1015706.1015755

Neil D. Lawrence, Matthias Seeger and Ralf Herbrich, Fast sparse Gaussian process methods: the informative vector machine,Proceedings of Neural Information Processing Systems 15. MIT Press:609-616,2003.

Martin F. Muller, A scaled conjugate gradient algorithm for fast supervised learning, Neural Networks,6(4):525-533,1993. http://dx.doi.org/10.1016/S0893-6080(05)80056-5



Most read articles by the same author(s)

Obs.: This plugin requires at least one statistics/report plugin to be enabled. If your statistics plugins provide more than one metric then please also select a main metric on the admin's site settings page and/or on the journal manager's settings pages.