Semi-Supervised Acoustic Model Training by Discriminative Data Selection From Multiple ASR Systems' Hypotheses.
Sheng LiYuya AkitaTatsuya KawaharaPublished in: IEEE ACM Trans. Audio Speech Lang. Process. (2016)
Keyphrases
- semi supervised
- computer systems
- data sets
- raw data
- data sources
- unlabeled data
- data analysis
- labeled data
- data processing
- database
- data structure
- labelled data
- original data
- data distribution
- synthetic data
- supervised learning
- support vector machine
- knowledge discovery
- data points
- end users
- training data
- neural network
- small number
- partially labeled
- labeled and unlabeled data
- feature selection
- data streams
- sensor data
- training set
- training samples
- semi supervised learning
- data collection
- image data
- input data