Decentralized, Unlabeled Multi-Agent Navigation in Obstacle-Rich Environments using Graph Neural Networks.
Xuebo JiHe LiZherong PanXifeng GaoChanghe TuPublished in: IROS (2021)
Keyphrases
- multi agent
- neural network
- cooperative
- multi agent environments
- mobile robot
- multi agent systems
- graph theory
- directed graph
- random walk
- multiagent systems
- topological map
- autonomous agents
- graph representation
- graph mining
- graph structure
- artificial neural networks
- pattern recognition
- semi supervised learning
- genetic algorithm
- active learning
- real world
- autonomous navigation
- indoor environments
- fuzzy logic
- structured data
- class labels
- connected components
- recurrent neural networks
- path planning
- labeled data
- graph model
- autonomous robots
- robot navigation
- multiple agents
- graph classification
- open systems
- navigation systems
- dynamic environments
- machine learning
- reinforcement learning
- unsupervised learning
- high level
- weighted graph
- multilayer perceptron
- neural network model
- website
- decision trees