Crowd-Aware Robot Navigation for Pedestrians with Multiple Collision Avoidance Strategies via Map-based Deep Reinforcement Learning.
Shunyi YaoGuangda ChenQuecheng QiuJun MaXiaoping ChenJianmin JiPublished in: IROS (2021)
Keyphrases
- robot navigation
- collision avoidance
- map building
- reinforcement learning
- continuous state
- topological map
- mobile robot
- autonomous mobile robot
- path planning
- autonomous robots
- landmark recognition
- visual navigation
- scene understanding
- dynamic environments
- optimal policy
- learning algorithm
- initially unknown
- indoor environments
- input output
- regression model
- high quality
- three dimensional
- pedestrian dynamics
- machine learning