Large margin training for hidden Markov models with partially observed states.
Trinh Minh Tri DoThierry ArtièresPublished in: ICML (2009)
Keyphrases
- hidden markov models
- partially observed
- discriminative training
- hidden states
- state transition
- baum welch
- speech recognition
- conditional random fields
- structured prediction
- minimum classification error
- markov models
- viterbi algorithm
- gesture recognition
- automatic speech recognition
- sequential data
- discriminative classifiers
- probabilistic interpretation
- image segmentation
- maximum likelihood
- acoustic models
- hierarchical hidden markov model
- markov model
- handwritten text recognition
- visual speech recognition
- higher order
- posterior probability
- sequence classification
- co training