ML parameter estimation of a multiscale stochastic process using the EM algorithm.
Ashvin KannanMari OstendorfWilliam Clement KarlDavid A. CastañónRandall K. FishPublished in: IEEE Trans. Signal Process. (2000)
Keyphrases
- parameter estimation
- stochastic process
- em algorithm
- multiscale
- maximum likelihood
- stochastic processes
- expectation maximization
- random fields
- markov chain
- stochastic model
- mixture model
- maximum likelihood estimation
- generative model
- hyperparameters
- wavelet transform
- image processing
- maximum a posteriori
- gibbs sampling
- image segmentation
- likelihood function
- hidden variables
- bayesian model selection
- wavelet domain
- approximate inference
- wavelet coefficients
- probability density function
- edge detection
- bayesian framework
- closed form
- higher order
- probability distribution
- hidden markov models
- computer vision