The Application of Hidden Markov Models in Speech RecognitionHidden Markov Models (HMMs) provide a simple and effective framework for modelling time-varying spectral vector sequences. As a consequence, almost all present day large vocabulary continuous speech recognition (LVCSR) systems are based on HMMs. Whereas the basic principles underlying HMM-based LVCSR are rather straightforward, the approximations and simplifying assumptions involved in a direct implementation of these principles would result in a system which has poor accuracy and unacceptable sensitivity to changes in operating environment. Thus, the practical application of HMMs in modern systems involves considerable sophistication. The Application of Hidden Markov Models in Speech Recognition presents the core architecture of a HMM-based LVCSR system and proceeds to describe the various refinements which are needed to achieve state-of-the-art performance. These refinements include feature projection, improved covariance modelling, discriminative parameter estimation, adaptation and normalisation, noise compensation and multi-pass system combination. It concludes with a case study of LVCSR for Broadcast News and Conversation transcription in order to illustrate the techniques described. The Application of Hidden Markov Models in Speech Recognition is an invaluable resource for anybody with an interest in speech recognition technology. |
Common terms and phrases
acoustic model adaptation data adaptive training algorithm approach approximation Architecture Audio Processing BN transcription cepstral clean speech cluster combination Computer Speech conditional independence confusion network continuous speech recognition covariance matrix criterion described discriminative training distribution dynamic Bayesian network example feature vector Gaussian component Gaussianisation graphemic hidden Markov models HLDA HMM-Based Recogniser ICSLP IEEE Transactions language model lattice linear transform M. J. F. Gales Mandarin maximise maximum a posteriori maximum likelihood maximum likelihood linear MLLR model parameters multi-pass multiple N-best N-gram noise robust normalisation output P. C. Woodland phone models posterior Proceedings of Eurospeech Proceedings of ICASSP pronunciations regression class robust speech recognition S. J. Young schemes segment signal processing speaker adaptation Speech and Audio Speech and Language speech recognition systems standard HMM techniques tion training data transcription system triphone uncertainty decoding unsupervised Viterbi Viterbi algorithm VTLN word sequence Wref