Self-supervised Deep Reinforcement Learning with Generalized Computation Graphs for Robot Navigation.
Gregory KahnAdam VillaflorBosen DingPieter AbbeelSergey LevinePublished in: CoRR (2017)
Keyphrases
- robot navigation
- reinforcement learning
- continuous state
- autonomous mobile robot
- autonomous robots
- scene understanding
- initially unknown
- real time stereo
- map building
- topological map
- function approximation
- graph matching
- object recognition
- model free
- action selection
- transfer learning
- optimal policy
- mobile robot
- support vector