The character model must be created in advance and given to the system. Because the character’s skeletal model is different from the actor’s skeletal model, captured motions and editing parameters are automatically retargetted from the user’s pose to the character’s pose.