Deep Energy: Using Energy Functions for Unsupervised Training of DNNs.
Alona GoltsDaniel FreedmanMichael EladPublished in: CoRR (2018)
Keyphrases
- energy minimization
- energy function
- graph cuts
- deep architectures
- markov random field
- supervised learning
- problems in computer vision
- global minimum
- image segmentation
- unsupervised learning
- active contours
- shape model
- belief propagation
- training set
- deep learning
- semi supervised
- graphical models
- segmentation method
- object segmentation
- energy functional
- conditional random fields
- shape prior
- approximate solutions
- bayesian networks
- machine learning