DEVIANT: Depth EquiVarIAnt NeTwork for Monocular 3D Object Detection.
Abhinav KumarGarrick BrazilEnrique CoronaArmin ParchamiXiaoming LiuPublished in: ECCV (9) (2022)
Keyphrases
- community structure
- d objects
- topological information
- pose estimation
- complex networks
- viewpoint
- object recognition
- multiple views
- multi view
- network structure
- three dimensional
- shape descriptors
- range data
- object features
- object detection
- depth map
- depth information
- depth cues
- cad models
- physically plausible
- line drawings
- image contours
- computer vision
- depth images
- three dimensional objects
- surface points
- pose parameters
- image sequences