A Tutorial on Hidden Markov Models and Selected Applications in Speech Recognition
Rabiner, 1989, A Tutorial on Hidden Markov Models and Selected Applications in Speech Recognition (Proceedings of the IEEE)
Subscriber deck — free during early access
2 cards in “IV Types of HMMs” · whole deck
IV Types of HMMs
- Contrast ergodic and left-right (Bakis) HMMs.Ergodic: every state reachable from every other (all \(a_{ij} > 0\) if fully connected). Left-right: the state index never decreases (\(a_{ij} = 0\) for \(j < i\)), starting in state 1 and ending in \(N\), often limiting jumps (\(a_{ij} = 0\) for \(j > i + \Delta\)); suited to signals whose properties change over time, like speech.
Left-right model with self-loops and skips of up to 2 states (Δ = 2) - Rabiner, 1989, A Tutorial on Hidden Markov Models and Selected Applications in Speech Recognition (Proceedings of the IEEE), IV Types of HMMs
- Why don’t left-right constraints complicate re-estimation?Any parameter initially set to zero stays zero throughout Baum–Welch re-estimation, so the topology is preserved automatically.
- Rabiner, 1989, A Tutorial on Hidden Markov Models and Selected Applications in Speech Recognition (Proceedings of the IEEE), IV Types of HMMs
Source
- Lawrence R. Rabiner. 1989. A Tutorial on Hidden Markov Models and Selected Applications in Speech Recognition. In Proceedings of the IEEE, pp. 257–286. doi:10.1109/5.18626