Motion Segmentation by Subspace Separation and Model Selection.
Ken-ichi KanataniPublished in: ICCV (2001)
Keyphrases
- motion segmentation
- model selection
- multibody
- cross validation
- hyperparameters
- low dimensional
- machine learning
- dimensionality reduction
- feature selection
- mixture model
- information criterion
- high dimensional data
- principal component analysis
- search space
- high dimensional
- feature space
- selection criterion
- model selection criteria
- unsupervised learning
- image sequences
- marginal likelihood
- supervised learning
- closed form
- bayesian information criterion
- image processing
- automatic model selection