MoGDE: Boosting Mobile Monocular 3D Object Detection with Ground Depth Estimation.
Yunsong ZhouQuan LiuHongzi ZhuYunzhe LiShan ChangMinyi GuoPublished in: CoRR (2023)
Keyphrases
- d objects
- depth estimation
- pose estimation
- depth map
- multi view
- viewpoint
- stereo vision
- object recognition
- three dimensional
- multiple views
- stereo matching
- range data
- image sequences
- depth information
- scene understanding
- dynamic scenes
- stereo pair
- semi supervised
- object classes
- position and orientation
- real scenes
- stereo camera
- multi class
- video sequences