Estimation of Cepstral Coefficients for Robust Speech Recognition

Acknowledgement Acknowledgement I would first like to thank my advisor, Dr. Richard Povinelli, whose advice and support have been critical in the development of this work. My thanks also go to Dr. Michael Johnson, whose expertise has been very valuable to me. I also thank my committee members, Drs. George Corliss, Craig Struble, and Edwin Yaz for all their assistance. Thanks go to Dr. Mohamed Mneimneh for all his help during his time as my labmate in the Knowledge and Information Discovery Lab. I am grateful to my family and all my friends for their moral support and friendship during my time as a student at Marquette University. Abstract Abstract This dissertation introduces a new approach to estimation of the features used in an automatic speech recognition system operating in noisy environments, namely mel-frequency cepstral coefficients. A major challenge in the development of an estimator for these features is the nonlinear interaction between a speech signal and the corrupting ambient noise. Previous estimation methods have attempted to deal with this issue with the use of a low order Taylor series expansion, which results in a rough approximation of the true distortion interaction between the speech and noise signal components, and the estimators must typically be iterative, as it is the speech features themselves that are used as expansion points. The new estimation approach, named the additive cepstral distortion model minimum mean-square error estimator, uses a novel distortion model to avoid the necessity of a Taylor series expansion, allowing for a direct solution. Like many previous approaches, the estimator introduced in this dissertation uses a prior distribution model of the speech features. In previous work, this distribution is limited in specificity, as a single global model is trained over an entire set of speech data. The estimation approach introduced in this work extends this method to incorporate contextual information into the prior model, leading to a more specific distribution and subsequently better estimates of the features.