Deep Optical Flow Estimation Via Multi-Scale Correspondence Structure Learning.
Shanshan ZhaoXi LiOmar El Farouk BourahlaPublished in: IJCAI (2017)
Keyphrases
- optical flow estimation
- structure learning
- multiscale
- graphical models
- bayesian networks
- optical flow
- coarse to fine
- feature matching
- conditional independence
- flow field
- parameter estimation
- motion model
- transfer learning
- stereo matching
- motion field
- markov networks
- image segmentation
- network structure
- sample size
- motion analysis
- velocity field
- computer vision
- point correspondences
- feature points
- probability distribution
- probabilistic model
- multi frame
- image processing
- optic flow
- object detection
- belief propagation