A Bayesian model of plan recognition
Artificial Intelligence
Using plan recognition in human-computer collaboration
UM '99 Proceedings of the seventh international conference on User modeling
Introduction to Reinforcement Learning
Introduction to Reinforcement Learning
Automated Assistants to Aid Humans in Understanding Team Behaviors
RoboCup-99: Robot Soccer World Cup III
| Hi-index | 0.01 |
We present an approach for automatically adapting the behavior of an autonomous table soccer robot to a human adversary. Basic actions are recognized as they are performed by the human player, and characteristic action observations are used to establish a model of the opponent. Based on this model, the opponent's playing skills are classified with respect to different levels of expertise and particular offensive and defensive skills are assessed. In response to the knowledge about the opponent, the robot adapts the velocities at which it attacks and defends in order to provide entertaining games for a wide range of human players with different playing skills. Experiments with two different table soccer robots validate our approach.