Retaining Informative Latent Variables in Probabilistic Segmentation.
M. M. Amaan ValiuddinChristiaan G. A. ViviersRuud van SlounPeter H. N. de WithFons van der SommenPublished in: ICASSP (2024)
Keyphrases
- latent variables
- probabilistic model
- latent variable models
- factor graphs
- conditionally independent
- real valued
- posterior distribution
- hierarchical model
- generative model
- approximate inference
- structural svms
- random variables
- image segmentation
- variational bayes
- topic models
- level set
- prior knowledge
- bayesian inference
- graphical models
- hidden variables
- structured output
- bayesian networks
- gaussian process
- language model
- max margin
- conditional random fields
- observed variables
- segmentation method
- causal relationships
- belief networks
- pairwise
- posterior probability
- conditional probabilities
- training data