Improved robustness of automatic speech recognition using a new class definition in linear discriminant analysis.
Martin SchaffönerMarcel KatzSven E. KrügerAndreas WendemuthPublished in: INTERSPEECH (2003)
Keyphrases
- linear discriminant analysis
- discriminant analysis
- scatter matrix
- fisher criterion
- class separability
- linear discriminant
- scatter matrices
- null space
- supervised dimensionality reduction
- face recognition
- dimensionality reduction
- principal component analysis
- discriminant features
- class discrimination
- dimension reduction
- support vector
- small sample size
- kernel discriminant analysis
- discriminant information
- feature space
- equivalence relationship
- linear transformation
- high dimensional data
- least squares
- feature extraction
- discriminative information
- support vector machine svm
- locality preserving projections
- pattern recognition
- computer vision
- subspace methods
- image processing
- qr decomposition
- speech signal
- optimization criterion
- high dimensionality
- facial expressions
- nearest neighbor
- feature selection