Continuous Control with Coarse-to-fine Reinforcement Learning.
Younggyo SeoJafar UruçStephen JamesPublished in: CoRR (2024)
Keyphrases
- coarse to fine
- reinforcement learning
- multiscale
- multiresolution
- object detection
- control problems
- dynamic programming
- image registration
- hierarchical segmentation
- optimal control
- matching scheme
- active shape model
- hierarchical representation
- action space
- image pyramid
- state space
- optical flow estimation
- action selection
- control policy
- scan line
- machine learning
- markov decision processes