Approximations of posterior distributions in blind deconvolution using variational methods.
Javier MateosRafael MolinaAggelos K. KatsaggelosPublished in: ICIP (2) (2005)
Keyphrases
- variational methods
- blind deconvolution
- posterior distribution
- approximate inference
- latent variables
- bayesian framework
- parameter estimation
- probability distribution
- optic flow
- hyperparameters
- maximum a posteriori
- optical flow
- level set
- gaussian process
- posterior probability
- optical flow estimation
- random variables
- curve evolution
- multi frame
- prior knowledge
- model selection
- probabilistic model
- generative model
- total variation
- missing values
- higher order
- markov random field
- image sequences
- prior information
- machine learning
- image processing
- computer vision