Automatic transcription of voicemail at AT&T

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

Volumn 1, Issue , 2001, Pages 25-28

  Bacchiani, M ^a  

^a AT AND T LABS RESEARCH   (United States)

Author keywords

[No Author keywords available]

Indexed keywords

CHANNEL NORMALIZATION; GENDER DEPENDENT SYSTEM; GENDER INDEPENDENT SYSTEM; LINEAR TRANSFORMATIONS; VOICEMAIL MESSAGES;

ALGORITHMS; COMMUNICATION CHANNELS (INFORMATION THEORY); CORRELATION METHODS; ERRORS; MATHEMATICAL TRANSFORMATIONS; PARAMETER ESTIMATION; SOUND RECORDING; SPEECH TRANSMISSION;

SPEECH RECOGNITION;

EID: 0034843020     PISSN: 15206149     EISSN: None     Source Type: Conference Proceeding    
DOI: None     Document Type: Conference Paper

References (9)

1. Vocal tract normalization in speech recognition: Compensating for systematic speaker variability
2. Speaker normalization on conversational telephone speech
3. Maximum likelihood linear regression for speaker adaptation of continuous density hidden Markov models
4. A compact model for speaker-adaptive training
5. Maximum likelihood linear transformations for HMM-based speech recognition
6. Semi-tied covariance matrices for hidden Markov models
7. Using maximum likelihood linear regression for segment clustering and speaker identification
8. Improving environmental robustness in large vocabulary speech recognition