LegoFormer: Transformers for Block-by-Block Multi-view 3D Reconstruction.
Farid YagubbayliAlessio TonioniFederico TombariPublished in: CoRR (2021)
Keyphrases
- block matching
- multi view
- block size
- motion vectors
- three dimensional
- optical flow
- single view
- multiple views
- depth map
- multi view stereo
- camera calibration
- visual hull
- range images
- d objects
- scene reconstruction
- stereo algorithms
- bundle adjustment
- multiple viewpoints
- multi view face detection
- semi supervised
- computer vision
- multi view clustering
- co training
- multiple cameras
- camera parameters
- depth information