PG-RCNN: Semantic Surface Point Generation for 3D Object Detection.
Inyong KooInyoung LeeSe-Ho KimHeeseon KimWoo-Jin JeonChangick KimPublished in: CoRR (2023)
Keyphrases
- d objects
- surface points
- object recognition
- viewpoint
- control points
- multi view
- shape descriptors
- range data
- pose estimation
- multiple views
- viewing direction
- three dimensional
- cad models
- pose normalization
- line drawings
- minimum bounding
- object detection
- texture mapping
- point distribution
- curved objects
- physically plausible
- d mesh
- real world objects
- surface features
- image contours
- surface patches
- free form
- object classes
- semantic information
- face recognition