Margin-based discriminant dimensionality reduction for visual recognition.
Hakan CevikalpBill TriggsFrédéric JurieRobi PolikarPublished in: CVPR (2008)
Keyphrases
- visual recognition
- dimensionality reduction
- principal component analysis
- low dimensional
- feature extraction
- locality preserving projections
- image classification
- linear dimensionality reduction
- discriminant analysis
- high dimensional data
- discriminant information
- multi class
- high dimensional
- linear discriminant analysis
- pattern recognition
- visual categorization
- visual recognition tasks
- data points
- object recognition
- manifold learning
- latent topic models
- view independent
- feature selection
- feature space
- unlabeled data
- face images
- generalization error
- feature representation
- pairwise constraints
- locally linear embedding
- metric learning
- category level
- image representation
- natural images
- feature vectors