Hide-and-Seek with the MOMDP Model
These videos show experiments with the Offline MOMDP model to play hide-and-seek, where the robot is the seeker and the person is the hider. Both players take one discrete step at the same time, making the time discrete.
Experiments
Triangle Reward
These experiments use the Triangle reward function to play hide-and-seek.
Simple Reward
These experiments use the Simple reward (a reward of 1 if winning and -1 when losing) to play hide-and-seek.