Multi-View Representation is What You Need for Point-Cloud Pre-Training.
Siming YanChen SongYoukang KongQixing HuangPublished in: ICLR (2024)
Keyphrases
- multi view
- point cloud
- multi view reconstruction
- multi view stereo
- single view
- surface reconstruction
- visual hull
- multiple views
- d objects
- depth map
- camera calibration
- bundle adjustment
- camera parameters
- semi supervised
- scene reconstruction
- laser scanner
- multiple cameras
- point sets
- photometric stereo
- structure from motion
- three dimensional
- multi view learning
- training set
- point correspondences
- range images
- training samples
- supervised learning
- high dimensional