Geometry-Aware Self-Training for Unsupervised Domain Adaptation on Object Point Clouds.
Longkun ZouHui TangKe ChenKui JiaPublished in: ICCV (2021)
Keyphrases
- point cloud
- laser scanner
- cad model
- real world objects
- structure from motion
- multi view stereo
- normal vectors
- point sets
- surface reconstruction
- triangular mesh
- stereo camera
- visual hull
- d objects
- point cloud data
- geometric information
- three dimensional
- training set
- multi view reconstruction
- object shape
- photorealistic
- relative position
- moving objects
- hough transform
- semi supervised learning
- geometric primitives