High degree-of-freedom hand model driven by lower degree-of-freedom input

This paper describes an estimation method to drive a hand model with high degree-of-freedom in a virtual reality space using lower degree-of-freedom hand posture input. Estimation of pose of each metacarpal bone inside the palm, which is not available in the lower degree-of-freedom hand posture input, is performed based on anatomical constraints between joints inside the hand. Such anatomical constraints are modeled as constraint functions of joint angles correlated to each other in hand motions.