A Maximum Likelihood Approach to Unsupervised Online Adaptation of Stochastic Vector Mapping Function for Robust Speech Recognition.
Donglai ZhuQiang HuoPublished in: ICASSP (4) (2007)
Keyphrases
- speech recognition
- mapping function
- noisy environments
- speaker adaptation
- speech recognizer
- language model
- hidden markov models
- speech synthesis
- pattern recognition
- speech signal
- automatic speech recognition
- speech recognition technology
- speaker independent
- unsupervised learning
- control points
- nonlinear dimensionality reduction
- speaker identification
- single channel
- semi supervised
- video sequences