MVContrast: Unsupervised Pretraining for Multi-view 3D Object Recognition.
Luequan WangHongbin XuWenxiong KangPublished in: Mach. Intell. Res. (2023)
Keyphrases
- multi view
- d objects
- range images
- multiple views
- semi supervised
- single view
- depth map
- multiple cameras
- object recognition
- multi view clustering
- three dimensional
- unsupervised learning
- co training
- scene reconstruction
- multi view images
- stereo video
- camera calibration
- surface reconstruction
- viewpoint
- pose estimation
- semi supervised learning
- supervised learning
- light field
- active learning
- multiple viewpoints
- multi view reconstruction
- d scene
- view synthesis
- depth video
- pairwise
- multi view face detection
- high quality