Toward Human-like Behavior Generation in Urban Environment Based on Markov Decision Process With Hybrid Potential Maps.
Chunzhao GuoKiyosumi KidonoRyuta TerashimaYoshiko KojimaPublished in: Intelligent Vehicles Symposium (2018)
Keyphrases
- urban environments
- markov decision process
- human behavior
- crowd behavior
- state space
- markov decision processes
- urban areas
- reinforcement learning
- optimal policy
- cad model
- augmented reality
- infinite horizon
- image processing
- reward function
- transition probabilities
- sufficient conditions
- learning algorithm
- route descriptions
- low rank textures