PFRL: Pose-Free Reinforcement Learning for 6D Pose Estimation.
Jianzhun ShaoYuhang JiangGu WangZhigang LiXiangyang JiPublished in: CoRR (2021)
Keyphrases
- pose estimation
- reinforcement learning
- human pose
- articulated objects
- object pose
- human body
- function approximation
- position and orientation
- d objects
- degrees of freedom
- human pose estimation
- computer vision
- head pose estimation
- machine learning
- feature points
- state space
- rigid objects
- learning algorithm
- optimal policy
- action selection
- dynamic programming
- learning process
- optimal control
- object recognition
- human hand
- markov decision processes
- learning classifier systems
- temporal difference
- viewpoint
- reinforcement learning algorithms
- body parts
- motion analysis
- policy search