PillarNeSt: Embracing Backbone Scaling and Pretraining for Pillar-based 3D Object Detection.
Weixin MaoTiancai WangDiankun ZhangJunjie YanOsamu YoshiePublished in: CoRR (2023)
Keyphrases
- d objects
- viewpoint
- object recognition
- multi view
- multiple views
- range data
- three dimensional
- pose estimation
- d mesh
- object detection
- object classes
- cad models
- shape descriptors
- detection method
- physically plausible
- pose normalization
- image contours
- object features
- control points
- three dimensional objects
- topological information
- object class
- line drawings