Playing Doom with SLAM-Augmented Deep Reinforcement Learning.
Shehroze BhattiAlban DesmaisonOndrej MiksikNantas NardelliN. SiddharthPhilip H. S. TorrPublished in: CoRR (2016)
Keyphrases
- reinforcement learning
- simultaneous localization and mapping
- mobile robot
- reinforcement learning algorithms
- state space
- computer games
- visual slam
- mobile robotics
- dynamic environments
- model free
- game playing
- function approximation
- particle filter
- machine learning
- transition model
- transfer learning
- temporal difference learning
- learning algorithm
- robotic control
- indoor environments
- kalman filter
- action selection
- optimal policy
- multi agent
- computer vision
- function approximators
- neural network
- loop closing