Convolutional neural network based deep conditional random fields for stereo matching.
Zhi WangShiqiang ZhuYuehua LiZhengzhe CuiPublished in: J. Vis. Commun. Image Represent. (2016)
Keyphrases
- conditional random fields
- stereo matching
- deep learning
- restricted boltzmann machine
- stereo vision
- belief propagation
- graphical models
- dynamic programming
- probabilistic model
- stereo images
- hidden markov models
- information extraction
- stereo correspondence
- higher order
- stereo matching algorithm
- markov random field
- disparity map
- depth map
- random fields
- post processing
- occlusion handling
- generative model
- pairwise
- image matching
- crf model
- sequence labeling
- data sets
- ground control points
- probabilistic graphical models
- decision trees
- information retrieval
- depth information
- multi class
- natural language
- reinforcement learning
- image sequences
- three dimensional
- machine learning