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