TC-Driver: A Trajectory Conditioned Reinforcement Learning Approach to Zero-Shot Autonomous Racing.
Edoardo GhignoneNicolas BaumannMichele MagnoPublished in: Field Robotics (2023)
Keyphrases
- reinforcement learning
- autonomous learning
- function approximation
- learning capabilities
- machine learning
- state space
- cooperative
- markov decision processes
- optimal policy
- reinforcement learning algorithms
- robotic systems
- learning process
- learning algorithm
- car navigation
- robotic control
- optimal control
- trajectory data
- artificial intelligence
- ifip th world computer congress
- transition model
- traffic accidents
- function approximators
- autonomous navigation
- supervised learning
- multi agent
- model free
- multi agent systems
- neural network
- spatial data
- dynamic programming