Leveraging Unlabeled Speech for Sequence Discriminative Training of Acoustic Models.
Ashtosh SapruSri GarimellaPublished in: INTERSPEECH (2020)
Keyphrases
- acoustic models
- discriminative training
- automatic speech recognition
- speech recognition
- hidden markov models
- maximum likelihood
- probabilistic interpretation
- posterior probability
- speech recognizer
- broadcast news
- co training
- support vector machine
- unlabeled data
- training data
- semi supervised learning
- speech signal
- unsupervised learning
- active learning
- speaker independent
- data sets
- labeled data
- generative model
- co occurrence
- supervised learning
- probability distribution
- neural network