Independent component analysis and synthesis of human motion

Editing human motion signals are difficult because of its multi-dimensional nature. In this paper, we propose a new method for analysis and synthesis of human motion using ICA. We extract several independent motion signals from human movement. The result of the extracted motion signal corresponds to the perceptually meaningful motion features that characterize the complex human actions. We show that generating various human motions can be much easier by editing motion feature extracted by ICA.