Mutual Information Divergence: A Unified Metric for Multimodal Generative Models.
Jin-Hwa KimYunji KimJiyoung LeeKang Min YooSang-Woo LeePublished in: CoRR (2022)
Keyphrases
- generative model
- mutual information
- multimodal image registration
- normalized mutual information
- probabilistic model
- information theoretic
- mixture model
- image registration
- similarity measure
- discriminative models
- discriminative learning
- conditional random fields
- prior knowledge
- semi supervised
- feature selection
- em algorithm
- information gain
- hierarchical hidden markov models
- expectation maximization
- maximum entropy principle
- object categories
- metric learning
- naive bayes models
- topic models
- maximum entropy
- learning process