SEFormer: Structure Embedding Transformer for 3D Object Detection.
Xiaoyu FengHeming DuYueqi DuanYongpan LiuHehe FanPublished in: CoRR (2022)
Keyphrases
- d objects
- viewpoint
- multi view
- pose estimation
- range data
- object recognition
- multiple views
- shape descriptors
- three dimensional
- object detection
- object classes
- object features
- d mesh
- physically plausible
- object views
- complex background
- object representation
- geometric models
- line drawings
- partial matching
- topological information
- cad models
- image sequences
- machine learning