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.D Explicit state duration density” · whole deck
IV.D Explicit state duration density
- What is the major weakness of conventional HMMs, according to Rabiner, and how can it be addressed?Modeling state duration: the implied duration density is geometric, inappropriate for most physical signals. A variable-duration (semi-Markov) HMM sets self-transitions to zero and draws an explicit duration \(p_i(d)\) for each state visit.
- Rabiner, 1989, A Tutorial on Hidden Markov Models and Selected Applications in Speech Recognition (Proceedings of the IEEE), IV.D Explicit state duration density
- What are the costs of explicit duration modeling?About \(D\) times the storage and \(D^2/2\) times the computation (a factor of about 300 for \(D = 25\)), many more parameters per state, and less data per parameter. Parametric durations (Gaussian, Gamma) or uniform ranges with constrained Viterbi help.
- Rabiner, 1989, A Tutorial on Hidden Markov Models and Selected Applications in Speech Recognition (Proceedings of the IEEE), IV.D Explicit state duration density
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