POMDPs for robotic tasks with mixed observability.
Sylvie C. W. OngShao Wei PngDavid HsuWee Sun LeePublished in: Robotics: Science and Systems (2009)
Keyphrases
- robotic tasks
- planning under uncertainty
- belief space
- partially observable
- partial observability
- partially observable markov decision processes
- motion planning
- belief state
- markov decision processes
- dynamic environments
- reinforcement learning
- state space
- decision problems
- decision theoretic
- robotic systems
- dynamic programming
- finite state
- ai planning
- dynamical systems
- multi agent
- point based value iteration
- planning problems
- path planning
- optimal policy
- degrees of freedom
- lower bound
- video sequences
- computer vision