Bayesian Inference of Models and Hyperparameters for Robust Optical-Flow Estimation.
Patrick HéasCédric HerzetÉtienne MéminPublished in: IEEE Trans. Image Process. (2012)
Keyphrases
- bayesian inference
- hyperparameters
- probabilistic model
- model selection
- optical flow estimation
- bayesian framework
- prior information
- variational bayes
- cross validation
- bayesian methods
- gaussian process
- gaussian processes
- posterior distribution
- optical flow
- random sampling
- incomplete data
- maximum a posteriori
- expectation propagation
- closed form
- parameter estimation
- em algorithm
- latent variables
- feature selection
- sample size
- graphical models
- markov random field
- prior knowledge
- support vector
- bayesian networks
- image sequences