Speech Recognition with Dynamic Bayesian Networks

  • Geoffrey Zweig

PhD Thesis: PhD Thesis University of California at Berkeley |

Dynamic Bayesian Networks (DBNs) are a powerful and flexible methodology for representing and computing with probabilistic models and stochastic processes. In the past decade, there has been increasing interest in applying them to practical problems, and this these shows that they can be used effectively in the field of automatic speech recognition.