Project: Designing Adaptive Team-mates for Games
MDA GAMBIT Grant, 2010-2011
Researcher(s): Kevin McGee
The goal of this project is to identify design knowledge for creating artificial team-mates that dynamically adapt to the needs and behaviors of players while cooperating to compete against adaptive AI opponents. Different phases of the research will explore the creation of partners for different genres of games; partners that cooperate in real-time with the player against one or more stationary and mobile opponents — and that do prediction, complex action-selection, and adaptation/learning.
Kevin McGee and Aswin Thomas Abraham. Real-time team-mate AI in Games: A Definition, Survey, & Critique. To appear in Proceedings of the 5th International Conference on Foundations of Digital Games, 19-21 June, 2010. Monterey, California, USA, 2010.
Aswin Thomas Abraham and Kevin McGee. AI for Dynamic Team-mate Adaptation in Games. To appear in IEEE Symposium on Computational Intelligence and Games, Copenhagen, Denmark, 18-21, August, 2010.