Feature extraction using pre-trained convolutive bottleneck nets for dysarthric speech recognition.
Yuki TakashimaToru NakashikaTetsuya TakiguchiYasuo ArikiPublished in: EUSIPCO (2015)
Keyphrases
- speech recognition
- pre trained
- feature extraction
- pattern recognition
- cepstral coefficients
- training data
- speaker identification
- training examples
- speech signal
- hidden markov models
- automatic speech recognition
- speech synthesis
- speech recognizer
- language model
- image processing
- feature vectors
- control signals
- image classification
- mel frequency cepstral coefficients
- speech recognition technology
- face recognition
- noisy environments
- neural network
- dimensionality reduction
- speech recognition systems
- speaker independent
- machine learning
- face images
- natural language processing
- feature space
- feature selection
- learning algorithm