SCPNet: Semantic Scene Completion on Point Cloud.
Zhaoyang XiaYouquan LiuXin LiXinge ZhuYuexin MaYikang LiYuenan HouYu QiaoPublished in: CVPR (2023)
Keyphrases
- point cloud
- laser scanner
- stereo camera
- urban scenes
- real world objects
- architectural scenes
- dominant plane
- surface reconstruction
- dense reconstruction
- structure from motion
- point sets
- multi view stereo
- d scene
- point cloud data
- cad model
- video sequences
- image sequences
- three dimensional
- photorealistic
- input image
- scene understanding
- single image
- moving objects
- ego motion
- iterative closest point
- multiple objects