Speaker adaptation of context dependent deep neural networks based on MAP-adaptation and GMM-derived feature processing.
Natalia A. TomashenkoYuri Y. KhokhlovPublished in: INTERSPEECH (2014)
Keyphrases
- context dependent
- speaker adaptation
- neural network
- speech recognition
- maximum likelihood
- gaussian mixture model
- speaker identification
- feature vectors
- context free
- speaker independent
- automatic speech recognition
- natural language
- pattern recognition
- low level
- speech recognizer
- high level
- speaker recognition
- product features
- image processing
- speaker dependent
- mixture model
- expectation maximization
- language model
- feature set
- video sequences
- information retrieval