ASD-Conv: Monocular 3D object detection network based on Asymmetrical Segmentation Depth-aware Convolution.
Xingyuan YuNeng DuGe GaoFan WenPublished in: ACML (2021)
Keyphrases
- d objects
- pose estimation
- multi view
- object recognition
- multiple views
- depth cues
- viewpoint
- depth map
- relative depth
- object contours
- three dimensional
- shape descriptors
- range data
- image sequences
- partial matching
- point distribution
- cad models
- object detection
- image segmentation
- depth images
- depth information
- level set
- stereo camera
- segmentation method
- segmentation algorithm
- image contours
- pose normalization
- object classes
- object features
- single view
- object segmentation
- surface patches
- bounding box
- topological information
- multiple objects
- edge detection
- video sequences
- physically plausible
- multi views