Cepstral parameter compensation for HMM recognition in noise
作者： M.J.F. Gales , S.J. Young
DOI: 10.1016/0167-6393(93)90093-Z
关键词: Bruit 、 Markov model 、 Speech processing 、 Speech recognition 、 Continuous density 、 Hidden Markov model 、 Cepstrum 、 Computer science 、 Robustness (computer science) 、 Compensation methods
摘要: Abstract This paper describes a method of adapting continuous density HMM recogniser trained on clean cepstral speech data to make it robust noise. The technique is based parallel model combination (PMC) in which the parameters corresponding pairs and noise states are combined yield set compensated parameters. It improves earlier mean compensation methods that also adapts variances as result can deal with much lower SNRs. PMC evaluated NOISEX-92 database shown work well down 0 dB SNR below for both stationary non-stationary noises. Furthermore, relatively constant conditions, there no additional computational cost at run-time.
elsevier.com
sci-hub.se 参考文章(15)
Steve J. Young, V. Beattie, Hidden Markov model state-based cepstral noise compensation. conference of the international speech communication association. ,(1992)
Andrew Varga, Keith Ponting, Control experiments on noise compensation in hidden Markov model based continuous word recognition. conference of the international speech communication association. pp. 1167- 1170 ,(1989)
D. Paul, A speaker-stress resistant HMM isolated word recognizer international conference on acoustics, speech, and signal processing. ,vol. 12, pp. 713- 716 ,(1987) , 10.1109/ICASSP.1987.1169551
J. Holmes, N. Sedgwick, Noise compensation for speech recognition using probabilistic models international conference on acoustics, speech, and signal processing. ,vol. 11, pp. 741- 744 ,(1986) , 10.1109/ICASSP.1986.1169209
D. Klatt, A digital filter bank for spectral matching international conference on acoustics, speech, and signal processing. ,vol. 1, pp. 573- 576 ,(1976) , 10.1109/ICASSP.1976.1170107
H.M. Cung, Y. Normandin, Noise adaptation algorithms for robust speech recognition Speech Communication. ,vol. 12, pp. 267- 276 ,(1993) , 10.1016/0167-6393(93)90098-6
B.H. Juang, Speech recognition in adverse environments Computer Speech & Language. ,vol. 5, pp. 275- 294 ,(1991) , 10.1016/0885-2308(91)90011-E
P. Lockwood, J. Boudy, Experiments with a Nonlinear Spectral Subtractor (NSS), Hidden Markov Models and the projection, for robust speech recognition in cars conference of the international speech communication association. ,vol. 11, pp. 215- 228 ,(1992) , 10.1016/0167-6393(92)90016-Z
H.B.D. Sorensen, A cepstral noise reduction multi-layer neural network international conference on acoustics, speech, and signal processing. pp. 933- 936 ,(1991) , 10.1109/ICASSP.1991.150493
D. Mansour, Biing Hwang Juang, A family of distortion measures base upon projection operation for robust speech recognition international conference on acoustics speech and signal processing. pp. 36- 39 ,(1988) , 10.1109/ICASSP.1988.196503