Robust Model Selection in 2D Parametric Motion Estimation.
Patrick BouthemyBertha Mayela Toledo AcostaBernard DelyonPublished in: J. Math. Imaging Vis. (2019)
Keyphrases
- model selection
- motion estimation
- cross validation
- motion segmentation
- hyperparameters
- sample size
- parameter estimation
- statistical learning
- bayesian learning
- regression model
- feature selection
- mixture model
- error estimation
- selection criterion
- model selection criteria
- statistical inference
- generalization error
- gaussian process
- meta learning
- super resolution
- variable selection
- optical flow
- image sequences
- video sequences
- machine learning
- bayesian information criterion
- information criterion
- marginal likelihood
- decision trees
- motion field
- bayesian methods
- support vector
- prior knowledge
- selection criteria
- mutual information
- parameter determination