Training data selection for improving discriminative training of acoustic models.
Berlin ChenShih-Hung LiuFang-Hui ChuPublished in: Pattern Recognit. Lett. (2009)
Keyphrases
- acoustic models
- discriminative training
- hidden markov models
- speech recognition
- probabilistic interpretation
- maximum likelihood
- automatic speech recognition
- training data
- support vector machine
- speech recognizer
- posterior probability
- broadcast news
- co training
- speaker independent
- em algorithm
- natural language processing
- decision trees