A visual interface for scripting virtual behaviors

The storyboarding stage in filmmaking is necessary for communicating overall scenarios rapidly through approximate visual drawings. In contrast, most 3D animation systems produce final products through repeated precise modeling. The interaction suggested in the paper promotes the maximum transfer of the users' knowledge of physical human actions into virtual environments, so the users can rapidly generate virtual human behaviors such as running, walking, grasping, and other motions. To achieve this, users visually describe the behaviors of a virtual human, then the interface parses the visual scripting and finally achieves the semantics. In this approach, users draw only curves, which describe synchronization among motions as well as geometric motion paths of the human, in the same three dimensional space using 3D devices such as a spaceball.