Image-based Regularization for Action Smoothness in Autonomous Miniature Racing Car with Deep Reinforcement Learning.
Hoang-Giang CaoI LeeBo-Jiun HsuZheng-Yi LeeYu-Wei ShihHsueh-Cheng WangI-Chen WuPublished in: IROS (2023)
Keyphrases
- reinforcement learning
- smoothing parameter
- action selection
- car racing
- autonomous learning
- partially observable domains
- prior information
- action space
- function approximation
- cooperative
- learning capabilities
- reward shaping
- fitted q iteration
- transition model
- state action
- state space
- cost function
- reinforcement learning algorithms
- computer vision
- temporal difference
- robotic systems
- regularization term
- regularization parameter
- markov decision process
- optimal control
- autonomous navigation
- autonomous mental development
- multi agent
- learning process
- evolutionary learning
- optimal policy
- supervised learning
- image segmentation
- model free
- dynamic programming
- agent learns
- markov decision processes
- regularization method
- function approximators
- smoothness constraint