Minimum classification error training of hidden Markov models for acoustic language identification.
Josef G. BauerEkaterina TimoshenkoPublished in: INTERSPEECH (2006)
Keyphrases
- minimum classification error
- hidden markov models
- language identification
- discriminative training
- statistical pattern recognition
- classification error
- maximum likelihood estimation
- document images
- speaker identification
- speech recognition
- conditional random fields
- gaussian mixture
- maximum likelihood
- baum welch
- speech signal
- support vector machine
- feature extraction
- probabilistic interpretation
- generative model
- markov random field