MeSA-DRL: Memory-Enhanced Deep Reinforcement Learning for Advanced Socially Aware Robot Navigation in Crowded Environments.
Mannan Saeed MuhammadEstrella MonteroPublished in: CoRR (2024)
Keyphrases
- robot navigation
- reinforcement learning
- autonomous robots
- continuous state
- autonomous mobile robot
- robot control
- multi agent
- mobile robot
- function approximation
- real time stereo
- scene understanding
- landmark recognition
- markov decision processes
- dynamic environments
- state space
- topological map
- map building
- learning algorithm
- initially unknown
- real environment
- object recognition
- robotic systems
- model free
- object detection
- machine learning
- real time