Defining an action of SO(d)-rotations on images generated by projections of d-dimensional objects: Applications to pose inference with Geometric VAEs.
Nicolas LegendreKhanh Dao DucNina MiolanePublished in: CoRR (2022)
Keyphrases
- polyhedral objects
- geometric information
- geometric constraints
- d objects
- three dimensional
- image regions
- partial occlusion
- individual objects
- input image
- image database
- multiple objects
- pose estimation
- cluttered background
- complex scenes
- real world objects
- object geometry
- bounding box
- object recognition and pose estimation
- target object
- image features
- object recognition
- rotation angles
- geometrical properties
- image retrieval
- geometric models
- multiple images
- geometric features
- image registration
- geometric structure
- object parts
- position and orientation
- complex background
- object representation
- keypoints
- computer vision
- wire frame
- visual input
- photometric properties
- object shape
- lighting conditions
- object models
- rigid body
- image set
- object class
- illumination conditions
- epipolar geometry