Introduction - If you have any usage issues, please Google them yourself
Hidden Markov Models (HMMs) are a popular approach
for speech recognition. Commonly, a left-to-right Markov
chain topology is used, where each phoneme is represented
by a sequence of states (typically three) [