Semi-Supervised Training of Acoustic Models Using Lattice-Free MMI.
Vimal ManoharHossein HadianDaniel PoveySanjeev KhudanpurPublished in: ICASSP (2018)
Keyphrases
- semi supervised
- acoustic models
- automatic speech recognition
- speech recognition
- supervised learning
- training process
- hidden markov models
- broadcast news
- labeled data
- semi supervised learning
- speech recognizer
- multi view
- unlabeled data
- discriminative training
- active learning
- training set
- speaker independent
- training examples
- generative model
- data sets
- multi modal
- natural language processing
- k means
- pairwise
- pattern recognition
- learning algorithm