Bayesian learning for hidden Markov model with Gaussian mixture state observation densities

Speech Communication

Volumn 11, Issue 2-3, 1992, Pages 205-213

  Gauvain, Jean Luc ^a   Lee, Chin Hui ^a  

^a LUCENT TECHNOLOGIES   (United States)

Author keywords

Bayesian learning; corrective training; Hidden Markov models; parameter smoothing; speaker adaptation; speaker clustering

Indexed keywords

LEARNING SYSTEMS; PROBABILITY;

BAYESIAN LEARNING; CORRECTIVE TRAINING; HIDDEN MARKOV MODEL; PARAMETER SMOOTHING; SPEAKER ADAPTATION; SPEAKER CLUSTERING;

SPEECH RECOGNITION;

EID: 0026882821     PISSN: 01676393     EISSN: None     Source Type: Journal    
DOI: 10.1016/0167-6393(92)90015-Y     Document Type: Article

References (15)

1. Bayesian adaptation in speech recognition
2. Maximum likelihood from incomplete data via the EM algorithm
3. Large-vocabulary speech recognition with speaker-adapted codebook and HMM parameters
4. Bayesian learning of Gaussian mixture densities for hidden Markov models
5. Maximum a posteriori estimation for multivariate Gaussian mixture observations of Markov chains
6. Improved hidden Markov modeling for speaker-independent continuous speech recognition
7. Interpolated estimation of Markov source parameters from sparse data
8. A study on speaker adaptation of continuous density HMM parameters
9. Acoustic modeling for large vocabulary speech recognition
10. Improved acoustic modeling for continuous speech recognition
11. An improed MMIE training algorithm for speaker-independent small vocabulary, continuous speech recognition
12. A segmental κ-means training procedure for connected word recognition
13. Dynamic speaker adaptation for feature-based isolated word recognition
14. A learning procedure for speaker-dependent word recognition systems based on sequential processing of input tokens