Unifying Voxel-based Representation with Transformer for 3D Object Detection.
Yanwei LiYilun ChenXiaojuan QiZeming LiJian SunJiaya JiaPublished in: CoRR (2022)
Keyphrases
- d objects
- invariant recognition
- multi view
- object recognition
- object representation
- viewpoint
- cad models
- three dimensional
- multiple views
- physically plausible
- topological information
- detection method
- pose estimation
- shape descriptors
- partial matching
- geometric models
- d mesh
- range data
- object classes
- pose normalization
- representation scheme
- control points
- object features
- image contours
- line drawings
- image representation
- visual hull
- object model
- structural information
- recognition and pose estimation