The 2001 BYBLOS English large vocabulary conversational speech recognition system

ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings

Volumn 1, Issue , 2002, Pages 721-724

  Matsoukas, Spyros ^a   Colthurst, Thomas ^a   Kimball, Owen ^a   Solomonoff, Alex ^a   Richardson, Fred ^a   Quillen, Carl ^a   Gish, Herbert ^a   Dognin, Pierre ^b  

^a BBN TECHNOLOGIES   (United States)

^b UNIVERSITY OF PITTSBURGH   (United States)

Author keywords

[No Author keywords available]

Indexed keywords

CEPSTRAL MEAN SUBTRACTION; GAUSSIAN SPLITTING INITIALIZATION; LINEAR DISCRIMINANT ANALYSIS; LINEAR PREDICTIVE CODING; PHONETICALLY TIED MIXTURE; SPEAKER ADAPTIVE TRAINING; STATE CLUSTERED TIED MIXTURE; VOCAL TRACT LENGTH NORMALIZATION;

ACOUSTIC SIGNAL PROCESSING; ALGORITHMS; DATA REDUCTION; DECODING; MARKOV PROCESSES; MATHEMATICAL MODELS; NATURAL LANGUAGE PROCESSING SYSTEMS; PARAMETER ESTIMATION; SPEECH ANALYSIS; SPEECH CODING; VOCABULARY CONTROL;

SPEECH RECOGNITION;

EID: 0036296880     PISSN: 15206149     EISSN: None     Source Type: Journal    
DOI: 10.1109/ICASSP.2002.5743819     Document Type: Article

References (10)

1. The 2000 BBN byblos LVCSR system
2. A compact model for speaker adaptive training
3. Maximum likelihood linear regression for speaker adaptation of continuous density HMMs
4. A post-processing system to yield reduced word error rates: Recognizer output voting error reduction (ROVER)
5. Experiments with a Gaussian merging-splitting algorithm for HMM training for speech recognition
6. Clustering via the bayesian information criterion with applications in speech recognition
7. Heteroscedastic discriminant analysis and reduced rank HMMs for improved speech recognition
8. Maximum likelihood discriminant feature spaces
9. Maximum likelihood modeling with Gaussian distributions for classification