SP-Net: A Sparse Convolution and Point-Encoding Enhanced Network for 3D Object Detection in LiDAR Point Clouds.
Meng LiuJianwei NiuYu LiuPublished in: RCAR (2023)
Keyphrases
- point cloud
- d objects
- lidar data
- real world objects
- cad model
- structure from motion
- multi view
- surface reconstruction
- object recognition
- viewpoint
- point cloud data
- pose estimation
- range data
- cad models
- normal vectors
- point sets
- stereo camera
- visual hull
- shape descriptors
- three dimensional
- laser scanner
- multiple views
- minimum bounding
- control points
- d mesh
- point distribution
- object classes
- surface patches
- high resolution