A minimum classification error based distance measure for template based speech recognition.
Mike MattonDirk Van CompernolleRonald CoolsPublished in: INTERSPEECH (2008)
Keyphrases
- distance measure
- speech recognition
- minimum classification error
- hidden markov models
- discriminative training
- statistical pattern recognition
- automatic speech recognition
- pattern recognition
- maximum likelihood estimation
- similarity measure
- speech synthesis
- language model
- distance function
- speech signal
- speech recognizer
- classification error
- speech recognition systems
- gaussian mixture
- speaker identification
- speaker independent
- em algorithm
- maximum likelihood
- probabilistic model
- training data
- information retrieval