Predictive minimum Bayes risk classification for robust speech recognition.
Jen-Tzung ChienKoichi ShinodaSadaoki FuruiPublished in: INTERSPEECH (2007)
Keyphrases
- speech recognition
- pattern recognition
- bayes risk
- noisy environments
- hidden markov models
- language model
- speech signal
- classification accuracy
- machine learning
- text classification
- speech recognition systems
- feature selection
- speaker identification
- support vector
- support vector machine
- feature vectors
- feature space
- loss function
- model selection
- similarity measure
- feature extraction
- computer vision