Descriptor-Free Multi-view Region Matching for Instance-Wise 3D Reconstruction.
Takuma DoiFumio OkuraToshiki NagaharaYasuyuki MatsushitaYasushi YagiPublished in: ACCV (5) (2020)
Keyphrases
- multi view
- region matching
- image retrieval
- three dimensional
- multi view stereo
- visual hull
- object matching
- single view
- d objects
- multiple views
- depth map
- shape descriptors
- object recognition
- camera calibration
- semi supervised
- computer vision
- structure from motion
- keypoints
- low level features
- visual features
- pairwise
- shape analysis
- action recognition
- geometric constraints
- multiscale
- feature space