MonoNeRD: NeRF-like Representations for Monocular 3D Object Detection.
Junkai XuLiang PengHaoran ChenHao LiWei QianKe LiWenxiao WangDeng CaiPublished in: ICCV (2023)
Keyphrases
- d objects
- pose estimation
- multi view
- object recognition
- viewpoint
- three dimensional
- shape descriptors
- detection method
- geometric models
- multiple views
- object classes
- cad models
- d mesh
- image sequences
- object representation
- line drawings
- range data
- partial matching
- topological information
- object detection
- object model
- representation scheme
- surface patches
- structural information
- real world objects
- hough transform
- surface points
- physically plausible
- recognition and pose estimation