Model-Based Approaches to Handling Uncertainty

作者: M. J. F. Gales

DOI: 10.1007/978-3-642-21317-5_5

关键词: Artificial intelligenceEstimation theoryMachine learningTask (project management)Uncertainty analysisAcoustic modelCompensation (engineering)NoiseTraining setStatistical modelComputer science

摘要: A powerful approach for handling uncertainty in observations is to modify the statistical model of data appropriately reflect this uncertainty. For task noise-robust speech recognition, requires modifying an underlying “clean” acoustic be representative a particular target environment. This chapter describes concepts model-based noise compensation robust recognition and how it can applied standard systems. The will then consider important practical issues. These include i) environment parameter estimation; ii) efficient likelihood calculation; iii) adaptive training handle multi-style data. conclude by discussing limitations current approaches research options address them.

参考文章(65)
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)
Dominique Fohr, Ángel de la Torre, Jean Paul Haton, Statistical adaptation of acoustic models to noise conditions for robust speech recognition. conference of the international speech communication association. pp. 1437- 1440 ,(2002)
Jon A. Arrowood, Mark A. Clements, Using observation uncertainty in HMM decoding. conference of the international speech communication association. ,(2002)
Alex Acero, Mike Plumpe, Li Deng, Xuedong Huang, Large-vocabulary speech recognition under adverse acoustic environments. conference of the international speech communication association. pp. 806- 809 ,(2000)
Alex Acero, Xuedong Huang, Hsiao-Wuen Hon, Spoken Language Processing Prentice-Hall. pp. 1008- ,(2001)
M. J. F. Gales, Model-based techniques for noise robust speech recognition Ph. D Dissertation, University of Cambridge. ,(1995)
Mark J. F. Gales, Rogier C. van Dalen, Asymptotically exact noise-corrupted speech likelihoods conference of the international speech communication association. pp. 709- 712 ,(2010)
H Liao, Mjf Gales, Joint uncertainty decoding for robust large vocabulary speech recognition University of Cambridge. ,(2006)
DK Kim, Mjf Gales, Adaptive training with noisy constrained maximum likelihood linear regression for noise robust speech recognition conference of the international speech communication association. pp. 2383- 2386 ,(2009)
Bhuvana Ramabhadran, Satya Dharanipragada, Ramesh A. Gopinath, Factor analysis invariant to linear transformations of data. conference of the international speech communication association. ,(1998)