Attention-based Depth Distillation with 3D-Aware Positional Encoding for Monocular 3D Object Detection.
Zizhang WuYunzhe WuJian PuXianzhi LiXiaoquan WangPublished in: CoRR (2022)
Keyphrases
- d objects
- pose estimation
- multi view
- multiple views
- depth cues
- viewpoint
- object recognition
- three dimensional
- depth images
- shape descriptors
- depth map
- range data
- d mesh
- image contours
- physically plausible
- three dimensional objects
- line drawings
- object classes
- partial matching
- topological information
- object detection
- image sequences
- stereo camera
- depth information
- human body
- surface points
- projective geometry
- vision system