Safer Autonomous Driving in a Stochastic, Partially-Observable Environment by Hierarchical Contingency Planning.
Ugo LecerfChristelle Yemdji TchassiPietro MichiardiPublished in: CoRR (2022)
Keyphrases
- partially observable
- autonomous driving
- partially observable markov decision process
- state space
- markov decision processes
- dynamical systems
- decision problems
- partial observability
- reinforcement learning
- fully observable
- infinite horizon
- markov decision problems
- belief state
- initially unknown
- planning problems
- belief space
- partial observations
- planning domains
- stereo vision
- dynamic environments
- dynamic programming
- real time
- mobile robot
- classical planning
- domain independent
- optimal policy
- grand challenge
- heuristic search
- knowledge base