Regularized constrained maximum likelihood linear regression for speech recognition.
Sina Hamidi GhalehjeghRichard C. RosePublished in: ICASSP (2014)
Keyphrases
- linear regression
- speech recognition
- maximum likelihood
- least squares
- speaker adaptation
- language model
- hidden markov models
- automatic speech recognition
- speech processing
- pattern recognition
- regression methods
- speech signal
- speech synthesis
- em algorithm
- speaker independent
- ridge regression
- speaker identification
- expectation maximization
- maximum a posteriori
- gaussian distribution
- mixture model
- speaker dependent
- noisy environments
- speech recognizer
- speech recognition technology
- speech recognizers
- isolated word
- signal processing