Module content
This module completes our coverage of the Hidden Markov Model, by looking at how a model for recognising connected speech can easily be built up from models of either words or sub-word units, in combination with a language model. To wrap up, we finally look at how HMMs are trained from data.
Here’s what you’re going to learn in this module:
Lecture slides
Thursday lecture slides (google) [updated 25/11/2024]