Improving eigenspace-based MLLR adaptation by kernel PCA.
Brian Kan-Wing MakRoger Wend-Huu HsiaoPublished in: INTERSPEECH (2004)
Keyphrases
- kernel pca
- principal component analysis
- kernel methods
- feature space
- laplacian eigenmaps
- speaker adaptation
- kernel function
- kernel principal component analysis
- spectral clustering
- face recognition
- input space
- gram matrix
- dimensionality reduction
- kernel matrix
- feature extraction
- component analysis
- computer vision
- support vector machine
- support vector
- input data
- low dimensional
- pattern recognition
- image segmentation
- data mining