About this document
HMMs in Speech Recognition Explained by Muni Sankar Matam is a document available to read on EtoBox.
This document discusses the use of hidden Markov models (HMMs) for speech recognition. It covers topics such as: - HMMs can model speech units like phones and words by representing speech as a sequence of observations. - Training HMM parameters involves using the forward-backward and Baum-Welch algorithms to optimize the model for observed speech data. - Decoding uses the Viterbi algorithm to find the most likely word sequence given an input speech signal and acoustic models.
- Author
- Muni Sankar Matam
- Language
- EN