TOMA: Topological Map Abstraction for Reinforcement Learning.
Zhao-Heng YinWu-Jun LiPublished in: CoRR (2020)
Keyphrases
- topological map
- reinforcement learning
- state abstraction
- robot navigation
- mobile robot
- indoor environments
- globally consistent
- topological features
- geometrical information
- robot localization
- state space
- autonomous navigation
- markov decision processes
- motion estimation
- simultaneous localization and mapping
- knowledge base
- machine learning
- loop closing
- map building
- general theory
- path planning
- logic programs
- image analysis
- learning algorithm