Towards non-stationary model-based noise adaptation for large vocabulary speech recognition

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

Volumn 1, Issue , 2001, Pages 337-340

  Kristjansson, T ^a   Frey, B ^a   Deng, L ^b   Acero, A ^b  

^a UNIVERSITY OF WATERLOO   (Canada)

^b MICROSOFT RESEARCH   (United States)

Author keywords

[No Author keywords available]

Indexed keywords

LARGE VOCABULARY SPEECH RECOGNITION; MEL FREQUENCY CEPSTRUM COMPONENT; NOISE ADAPTATION; VECTOR SERIES METHOD;

MATHEMATICAL MODELS; MATHEMATICAL TRANSFORMATIONS; SPEECH ANALYSIS; VECTORS;

SPEECH RECOGNITION;

EID: 0034849087     PISSN: 15206149     EISSN: None     Source Type: Journal    
DOI: 10.1109/ICASSP.2001.940836     Document Type: Article

References (10)

1. Hidden markov model decomposition of speech and noise
2. Speech recognition with dynamic baysian networks
3. Statistical-model-based speech enhancement systems
4. Hmm adaptation using vector taylor series for noisy speech recognition
5. Speech recognition in adverse environments
6. Suppression of acoustic noise in speech using spectral subtraction