Issues with Uncertainty Decoding for Noise Robust Speech Recognition
作者： M. J. F. Gales , Hank Liao
DOI:
关键词:
摘要: Recently there has been interest in uncertainty decoding for robust speech recognition. Here the associated with observation noise is propagated to recogniser. By using appropriate approximations this uncertainty, it possible obtain efficient implementations during decoding. The aim of these schemes performance which close that a modelbased compensated system, without computational cost. Unfortunately, low SNR isa fundamental issue withfront-end where model means and variances are updated according features. This described detail Joint SPLICE forms, but not limited two techniques. A solution scheme presented along implicit approach used uncertainty. In addition, model-based described, more powerful than front-end schemes, being affected by problem. illustrated AURORA 2.0 database various systems. Index Terms: compensation, recognition,
isca-speech.org参考文章(7)
Hugo Van hamme, Veronique Stouten, Patrick Wambacq, Accounting for the uncertainty of speech estimates in the context of model-based feature enhancement international conference on spoken language processing. pp. 105- 108 ,(2004)
Li Deng, Jasha Droppo, Alex Acero, Exploiting variances in robust feature extraction based on a parametric model of speech distortion. conference of the international speech communication association. ,(2002)
M.J.F. Gales, Maximum likelihood linear transformations for HMM-based speech recognition Computer Speech & Language. ,vol. 12, pp. 75- 98 ,(1998) , 10.1006/CSLA.1998.0043
M.J.F. Gales, Predictive model-based compensation schemes for robust speech recognition Speech Communication. ,vol. 25, pp. 49- 74 ,(1998) , 10.1016/S0167-6393(98)00029-6
Jasha Droppo, Alex Acero, Li Deng, Uncertainty decoding with SPLICE for noise robust speech recognition international conference on acoustics, speech, and signal processing. ,vol. 1, pp. 57- 60 ,(2002) , 10.1109/ICASSP.2002.5743653
M. J. F. Gales, Hank Liao, Joint uncertainty decoding for noise robust speech recognition conference of the international speech communication association. pp. 3129- 3132 ,(2005)
Trausti T. Kristjansson, Brendan J. Frey, Accounting for uncertainity in observations: A new paradigm for Robust Automatic Speech Recognition international conference on acoustics, speech, and signal processing. ,vol. 1, pp. 61- 64 ,(2002) , 10.1109/ICASSP.2002.5743654