Nonlinear Dimensional Reduction and Generation of Human Motion

To reuse existing motion data to generate new motions, a nonlinear dimensional reduction and generation of human motion is presented, which is based on fast adaptive scaled Gaussian process latent variable models.Through statistical learning, the motion data are mapped from high-dimensional observation space to low-dimensional latent space to implement nonlinear dimensional 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 speed and precision and extends editing range of motion by adapting motion editing direction.