Use of generalized dynamic feature parameters for speech recognition: maximum likelihood and minimum classification error approaches.
Rathinavelu ChengalvarayanLi DengPublished in: ICASSP (1995)
Keyphrases
- speech recognition
- maximum likelihood
- minimum classification error
- hidden markov models
- maximum likelihood estimation
- speaker adaptation
- speech recognition systems
- discriminative training
- speech synthesis
- expectation maximization
- parameter estimation
- speech signal
- em algorithm
- speech recognizer
- pattern recognition
- statistical pattern recognition
- automatic speech recognition
- speaker identification
- gaussian mixture
- language model
- noisy environments
- mixture model
- cepstral coefficients
- maximum a posteriori
- speaker independent
- speech recognition technology