PFRL: Pose-Free Reinforcement Learning for 6D Pose Estimation.
Jianzhun ShaoYuhang JiangGu WangZhigang LiXiangyang JiPublished in: CVPR (2020)
Keyphrases
- pose estimation
- reinforcement learning
- human pose
- articulated objects
- function approximation
- human body
- d objects
- state space
- object pose
- head pose estimation
- learning algorithm
- computer vision
- markov decision processes
- rigid objects
- human pose estimation
- body parts
- multi agent
- reinforcement learning algorithms
- model free
- human hand
- robotic control
- position and orientation
- feature points
- action selection
- human detection
- depth images
- optimal control
- action space
- object tracking
- action recognition
- dynamic programming
- three dimensional