A New Approach Combining Trained Single-view Networks with Multi-view Constraints for Robust Multi-view Object Detection and Labelling.
Yue ZhangAdrian HiltonJean-Yves GuillemautPublished in: VISIGRAPP (5: VISAPP) (2020)
Keyphrases
- multi view
- single view
- object detection
- multiple views
- view invariant
- perspective projection
- depth map
- d objects
- three dimensional
- co training
- multiple cameras
- semi supervised
- camera calibration
- photometric stereo
- bundle adjustment
- camera parameters
- vanishing points
- scene reconstruction
- aam fitting
- multi view clustering
- computer vision
- surface reconstruction
- range images
- training set
- multiple viewpoints
- multi view learning
- surface normals
- background subtraction
- stereo matching
- face detection
- object recognition