Decentralized Structural-RNN for Robot Crowd Navigation with Deep Reinforcement Learning.
Shuijing LiuPeixin ChangWeihang LiangNeeloy ChakrabortyKatherine Driggs CampbellPublished in: CoRR (2020)
Keyphrases
- reinforcement learning
- obstacle avoidance
- mobile robot
- robot navigation
- multi robot systems
- robot control
- multi agent
- rough terrain
- real robot
- indoor environments
- recurrent neural networks
- unknown environments
- service robots
- autonomous navigation
- navigation tasks
- path planning
- autonomous learning
- perceptual aliasing
- multi robot
- visual navigation
- autonomous vehicles
- autonomous robots
- nearest neighbor
- potential field
- cooperative
- collision free
- outdoor environments
- function approximation
- structural information
- collision avoidance
- markov decision processes
- vision system
- state space
- learning process
- mobile robotics
- learning algorithm
- human robot interaction
- humanoid robot
- robot behavior
- robotic control
- decentralized control
- dead reckoning
- multiagent reinforcement learning
- multi robot coordination
- robotic arm
- reinforcement learning algorithms
- model free
- position and orientation
- motion planning
- robotic systems
- optimal policy
- peer to peer
- sensory inputs
- robot manipulators
- topological map
- simultaneous localization and mapping
- temporal difference
- optimal control
- distributed systems
- dynamic programming
- genetic algorithm