Probabilistic linear discriminant analysis with bottleneck features for speech recognition.
Liang LuSteve RenalsPublished in: INTERSPEECH (2014)
Keyphrases
- speech recognition
- linear discriminant analysis
- feature extraction
- feature space
- discriminant features
- fisher criterion
- speech recognition systems
- discriminant analysis
- hidden markov models
- discriminative information
- speech signal
- principal component analysis
- pattern recognition
- dimension reduction
- face recognition
- dimensionality reduction
- speech recognizer
- class separability
- feature set
- speaker identification
- cepstral coefficients
- null space
- speech synthesis
- generative model
- noisy environments
- automatic speech recognition
- language model
- probabilistic model
- feature vectors
- data sets
- feature selection
- information retrieval
- speech recognition technology
- high dimensional data
- image features
- support vector
- machine learning