EARL: Eye-on-Hand Reinforcement Learner for Dynamic Grasping with Active Pose Estimation.
Baichuan HuangJingjin YuSiddarth JainPublished in: CoRR (2023)
Keyphrases
- pose estimation
- hand pose
- pose recovery
- human body
- degrees of freedom
- computer vision
- human pose estimation
- d objects
- monocular images
- problems in computer vision
- head pose estimation
- pose parameters
- articulated model
- position and orientation
- depth images
- human hand
- body parts
- human pose
- vision system
- feature points
- camera pose
- shape model
- eye movements
- action recognition
- viewpoint
- video sequences
- reinforcement learning