VI-Net: Boosting Category-level 6D Object Pose Estimation via Learning Decoupled Rotations on the Spherical Representations.
Jiehong LinZewei WeiYabin ZhangKui JiaPublished in: ICCV (2023)
Keyphrases
- pose estimation
- d objects
- viewpoint
- computer vision
- learning algorithm
- human body
- pose parameters
- three dimensional
- object recognition
- feature points
- multi view
- active learning
- multiple views
- shape descriptors
- visual recognition
- training data
- single image
- decision trees
- object classes
- image descriptors
- feature selection
- invariant recognition