Algorithm for real-time style transfer for human motion

Motion capture has become a very important technique in computer animation. Because of the high cost of motion data, researchers have focused their research work on how to edit the captured motion data, and how to efficiently synthesize the new motions based on available motion database. Comparing to existing research work on motion editing and synthesis, relatively few work was done on motion style editing. The problem of motion style transfer was expressed as a problem of motion statistic transfer, and a fast and convenient algorithm for human-motion style transfer was proposed. Experiment results demonstrate that the approach has the advantages of fast execution, low memory occupation, and easy implementation. The proposed algorithm can be widely applied to computer animation, game and digital film production.