Causal and semicausal AR image model identification using the EM algorithm.
Yücel YemezEmin AnarimYorgo IstefanopulosPublished in: IEEE Trans. Image Process. (1993)
Keyphrases
- em algorithm
- parameter estimation
- gaussian mixture
- expectation maximization
- probability density function
- maximum likelihood
- statistical model
- mixture distribution
- maximum likelihood estimation
- multiscale
- maximum likelihood estimates
- gibbs sampling
- gaussian mixture model
- image segmentation
- map estimation
- log likelihood
- parameter space
- closed form
- mixture model
- image features
- probabilistic model
- prior information
- energy function
- bayesian framework
- generative model
- hidden variables
- model based clustering
- gaussian distribution
- incomplete data
- normal distribution
- feature vectors
- bayesian networks
- similarity measure
- image sequences