Variational-Bayes Optical Flow.
Giannis K. ChantasTheodosios GkamasChristophoros NikouPublished in: J. Math. Imaging Vis. (2014)
Keyphrases
- variational bayes
- optical flow
- bayesian inference
- latent variables
- bayesian learning
- gaussian mixture model
- hyperparameters
- posterior distribution
- free energy
- image sequences
- optical flow estimation
- motion model
- latent dirichlet allocation
- motion field
- exponential family
- motion estimation
- flow field
- least squares
- computer vision
- probabilistic model
- moving objects
- cross validation
- incremental learning
- maximum likelihood
- expectation maximization
- log likelihood
- noise reduction
- bayesian framework
- prior information
- random sampling
- closed form
- prior knowledge
- parameter estimation