Active learning workflows and integrable deep neural networks for representing the free energy functions of alloy.
Gregory H. TeichertAnirudh NatarajanAnton Van der VenKrishna C. GarikipatiPublished in: CoRR (2020)
Keyphrases
- energy function
- active learning
- neural network
- graph cuts
- energy minimization
- gradient field
- markov random field
- problems in computer vision
- semi supervised
- active contours
- belief propagation
- global minimum
- supervised learning
- level set
- training set
- image segmentation
- shape model
- approximate solutions
- energy functional
- shape prior
- machine learning
- training examples
- information theoretic
- smoothness constraint
- genetic algorithm
- greedy algorithm
- segmentation method
- unlabeled data
- competitive learning
- input image
- learning algorithm
- regularization framework
- submodular functions