Flexible Feature Spaces Based on Generalized Heteroscedastic Linear Discriminant Analysis.
Alessandro DuminucoChaojun LiDavid KryzeLuca RigazioPublished in: ICASSP (1) (2006)
Keyphrases
- linear discriminant analysis
- feature space
- dimensionality reduction
- discriminant analysis
- principal component analysis
- small sample size
- dimension reduction
- feature extraction
- face recognition
- high dimensional
- feature selection
- discriminant features
- null space
- high dimensional data
- fisher criterion
- kernel function
- input space
- high dimensionality
- support vector
- training set
- low dimensional
- principal components
- class separability
- feature vectors
- training samples
- face detection
- linear discriminant
- pattern recognition
- high dimensional feature space
- support vector machine
- hyperplane
- multiple features
- subspace methods
- discriminative information
- computer vision
- kernel methods
- support vector machine svm
- input image
- lower dimensional
- machine learning
- feature set
- data points
- covariance matrix