NavRep: Unsupervised Representations for Reinforcement Learning of Robot Navigation in Dynamic Human Environments.
Daniel DugasJuan I. NietoRoland SiegwartJen Jen ChungPublished in: ICRA (2021)
Keyphrases
- robot navigation
- autonomous robots
- reinforcement learning
- continuous state
- autonomous mobile robot
- dynamic environments
- scene understanding
- supervised learning
- robot control
- unsupervised learning
- map building
- real time stereo
- initially unknown
- mobile robot
- function approximation
- dynamic programming
- image sequences
- model free
- object recognition
- single image