2D3D-MVPNet: Learning cross-domain feature descriptors for 2D-3D matching based on multi-view projections of point clouds.
Baiqi LaiWeiquan LiuCheng WangXiaoliang FanYangbin LinXuesheng BianShangbin WuMing ChengJonathan LiPublished in: Appl. Intell. (2022)
Keyphrases
- multi view
- point cloud
- cross domain
- multi view reconstruction
- feature descriptors
- single view
- multi view stereo
- surface reconstruction
- learning algorithm
- knowledge transfer
- three dimensional
- learning process
- co training
- multiple views
- learning tasks
- semi supervised
- pose estimation
- prior knowledge
- shape descriptors
- supervised learning
- image registration
- d objects
- computer vision
- unsupervised learning
- feature points
- image matching
- image features
- object recognition
- reinforcement learning
- training data
- machine learning