Sim-to-Real Transfer with Incremental Environment Complexity for Reinforcement Learning of Depth-Based Robot Navigation.
Thomas ChaffreJulien MorasAdrien Chan-Hon-TongJulien MarzatPublished in: CoRR (2020)
Keyphrases
- robot navigation
- autonomous mobile robot
- autonomous robots
- reinforcement learning
- continuous state
- initially unknown
- real time stereo
- map building
- topological map
- mobile robot
- landmark recognition
- real environment
- robot localization
- real robot
- scene understanding
- object detection
- decision problems
- transfer learning
- real time
- dynamic programming
- depth information
- model free
- video surveillance
- event detection
- optimal policy
- pairwise
- object recognition
- high quality
- machine learning