Interpolation Synthesis of Articulated Figure Motion

ost conventional media depend on engaging and appealing characters. Empty spaces and buildings would not fare well as television or movie programming, yet virtual reality usually offers up such spaces. The problem lies in the difficulty of creating computer-generated characters that display real-time, engaging interaction and realistic motion. Articulated figure motion for real-time computer graphics offers one solution to this problem. A common approach stores a set of motions and lets you choose one particular motion at a time. This article describes a process that greatly expands the range of possible motions. Mixing motions selected from a database lets you create a new motion to exact specifications. The synthesized motion retains the original motions' subtle qualities, such as the realism of motion capture or the expressive, exaggerated qualities of artistic animation. Our method provides a new way to achieve inverse kine-matics capability—for example, placing the hands or feet of an articulated figure in specific positions. It proves useful for both real-time graphics and prerendered animation production. Video production in computer-based and traditional animation relies largely on the labor-intensive process of keyframing. 1 This does not permit creating real-time characters with intriguingly rich varieties of motion. Motion capture allows more rapid production of character motion, but still suffers from the inherent lack of control of prerecorded motion data. Physical simulation approaches offer promise, but are hampered by lack of control, difficulty of use, instabilities, and computational cost that usually precludes real-time operation. 2 The predominant approach to real-time motion synthesis , selection of one stored motion at runtime, restricts computation to the stored motion's display. Unfortunately, a limited number of stored motions can result in repetitive or incorrect motions. The repertoire of possible motions expands greatly— at the cost of additional computation—with interpolation synthesis. This technique combines a set of motions similar to that desired to form an exactly specified motion. A small set of example motions then yields a continuous multidimensional space of possible motions. The example motions can come from keyframing, physical simulations, or motion capture. Because the interpolation process generally preserves the motion's subtle qualities, we can create reusable libraries of motion data from a single step of laborious keyframing, iterative estimation of initial conditions, or correction of errors and dropouts, respectively. This article focuses on real-time interpolation synthesis of motion based on motion capture data. This approach works well for real-time character motion in interactive entertainment or …