An Unsupervised Learning Approach to Discontinuity-Preserving Image Registration.
Eric NgMehran EbrahimiPublished in: WBIR (2020)
Keyphrases
- discontinuity preserving
- unsupervised learning
- image registration
- surface reconstruction
- optical flow estimation
- graph cuts
- motion field
- energy minimization
- coarse to fine
- smoothness constraint
- map estimation
- maximum a posteriori
- nonlinear diffusion
- markov random field
- mutual information
- image pairs
- deformation field
- phase unwrapping
- expectation maximization
- image matching
- stereo correspondence
- optical flow
- semi supervised
- dimensionality reduction
- machine learning
- medical imaging
- image intensity
- feature matching
- image analysis
- point cloud
- object recognition
- maximum likelihood
- motion model
- motion estimation
- energy function
- model selection
- em algorithm
- feature selection
- image denoising
- three dimensional
- image sequences
- multiresolution
- range images
- belief propagation