MSDA: Monocular Self-supervised Domain Adaptation for 6D Object Pose Estimation.
Dingding CaiJanne HeikkiläEsa RahtuPublished in: SCIA (2) (2023)
Keyphrases
- pose estimation
- domain adaptation
- d objects
- viewpoint
- multi view
- object recognition
- human body
- multiple views
- three dimensional
- range data
- transfer learning
- human pose
- monocular images
- semi supervised
- labeled data
- pose parameters
- shape descriptors
- position and orientation
- multiple sources
- semi supervised learning
- body parts
- depth images
- invariant recognition
- computer vision
- recognition and pose estimation
- object classes
- projective geometry
- point correspondences
- prior knowledge