ST3D++: Denoised Self-training for Unsupervised Domain Adaptation on 3D Object Detection.
Jihan YangShaoshuai ShiZhe WangHongsheng LiXiaojuan QiPublished in: CoRR (2021)
Keyphrases
- d objects
- multi view
- viewpoint
- range data
- object recognition
- cad models
- object detection
- shape descriptors
- line drawings
- pose estimation
- three dimensional
- denoising
- detection method
- multiple views
- d mesh
- physically plausible
- object classes
- three dimensional objects
- object representation
- object features
- surface points
- image contours
- geometric invariants
- single view
- control points
- surface patches
- topological information
- machine learning