OTPose: Occlusion-Aware Transformer for Pose Estimation in Sparsely-Labeled Videos.
Kyung-Min JinGun-Hee LeeSeong-Whan LeePublished in: CoRR (2022)
Keyphrases
- pose estimation
- human pose
- static images
- human body
- computer vision
- d objects
- body parts
- monocular images
- human pose estimation
- video sequences
- human actions
- position and orientation
- head pose estimation
- pose parameters
- depth images
- action recognition
- feature points
- video frames
- pictorial structure
- training data
- video surveillance
- training set
- partial occlusion
- human motion
- human activities
- object pose
- background clutter
- viewpoint
- object recognition and pose estimation