Dimensionality reduction for speech emotion features by multiscale kernels.
Xinzhou XuJun DengWenming ZhengLi ZhaoBjörn W. SchullerPublished in: INTERSPEECH (2015)
Keyphrases
- multiscale
- dimensionality reduction
- feature extraction
- feature space
- high dimensionality
- low level
- image processing
- structural features
- translation invariant
- kernel learning
- feature set
- emotion classification
- neural network
- tree kernels
- principal components
- scale space
- feature vectors
- image representation
- text categorization
- linear discriminant analysis
- principal component analysis
- coarse to fine
- data representation
- support vector machine
- emotional state
- dimensionality reduction methods
- high dimensional
- nonlinear dimensionality reduction
- pattern recognition
- lexical features
- face recognition