Dynamically Simulated Characters in Virtual Environments
Dynamically Simulated Characters in Virtual Environments home page
Movies from our 1997 Siggraph Animation Sketch
More about our current virtual environment hardware setup
Slide 2 of 21
There are many potential domains for virtual environments, but we are interested in those that possess compelling multiagent interactions. On this slide I have placed photos from a couple real-world scenarios that exemplify these kinds of interactions.
The bicyclists in the top photo are racers from the 1996 Olympics and demonstrate the uses of virtual environments in sports training. Each bicyclist must maintain balance, stay on the road, decide when to push hard and when to stay back all while in close proximity to other cyclists. The relationship between the dynamic properties of each cyclist and their higher-level behaviors must be carefully tuned so that each cyclist capitalizes on his or her strength. Users can use our system to experience riding through a course with teammates and opponents. The user will learn about the geographic contours and landmarks of the course and will learn how to ride with others as well.
The border collie in the bottom photo exemplifies an entertaining virtual environment application. This dog has been bread to possess the innate skill to use its physical position and velocity to move a flock of sheep. A human might not find this task intuitive, but he or she can play the role of the dog when trying to herd one-legged robots into a corral. A final application area for dynamic simulations in multiagent virtual environments is that of task training. Professionals such as firemen and policemen must often execute finely tuned coordinated movements to complete a task. Multiagent virtual environments allow these people to train without placing themselves in danger or at great expense.
Slide 2 of 21