CrossFusion net: Deep 3D object detection based on RGB images and point clouds in autonomous driving.
Dza-Shiang HongHung-Hao ChenPei-Yung HsiaoLi-Chen FuSiang-Min SiaoPublished in: Image Vis. Comput. (2020)
Keyphrases
- d objects
- point cloud
- real world objects
- object recognition
- integral imaging
- three dimensional
- cad model
- multi views
- image data
- surface reconstruction
- multi view
- surface properties
- shape descriptors
- autonomous driving
- image retrieval
- structure from motion
- range data
- input image
- object matching
- viewpoint
- image matching
- pose estimation
- image features
- multiple views
- multiple images
- surface patches
- traffic signs
- keypoints
- object detection
- point sets
- range images
- computer aided design
- viewing direction
- feature points
- training set
- machine learning