Demos and Descriptions of Projects
Many of our projects, because they involve three-dimensional virtual environments, also allow us to show you what our software does. Other projects benefit from a description that's heavy on the screen shots and light on the citations (see the publications page for those in-depth explanations). The demos and descriptive pages linked below include brief explanations of what they are and a chance to see our work in action for yourself.
Rapid Tactics Generation Phase I and Phase II
What if, instead of programming behavior to create an intelligent
agent, you could just demonstrate it? This project, a Phase I
SBIR sponsored by NAVAIR, aims to accomplish exactly that. We
have targeted two virtual environments. The first environment,
the dTank tank simulator, is a simple two-dimensional environment that
has allowed us to rapidly prototype and demonstrate behaviors.
The second environment, Unreal Tournament, is a full-blown immersive
three-dimensional game. The principles are the same, but the
graphics are certainly more compelling.
Adaptive Mesh Refinement (AMR)
One of the challenges in Cognitive Modeling is navigating the enormous space of possibilities in configuring intelligent agents, or even in conducting basic research. How many variations on agent behavior would you like to test? Is there time to do so? This project focuses on optimizing data collection for intelligent architectures so the modeler or the scientist can spend their computational resources where they make a difference, without squandering money or time where they don't.
Training with Asymmetric Opponents (TAO)
Opponents confronted with superior opponents or uneven numbers often resort to unusual tactics in an effort to even the odds. This project focuses on modeling the kinds of reasoning that take place in exactly these uneven situations, which are more the rule than the exception in many domains, and embedding entities using asymmetric reasoning within virtual environments.