Semi-Supervised End-to-End ASR via Teacher-Student Learning with Conditional Posterior Distribution.
Zi-qiang ZhangYan SongJian-Shu ZhangIan McLoughlinLi-Rong DaiPublished in: INTERSPEECH (2020)
Keyphrases
- end to end
- student learning
- posterior distribution
- semi supervised
- student interaction
- probability distribution
- latent variables
- parameter estimation
- formative assessment
- student achievement
- bayesian framework
- hyperparameters
- learning tools
- collaborative learning
- posterior probability
- maximum a posteriori
- conditional probabilities
- gaussian distribution
- instructional practices
- generative model
- active learning
- labeled data
- learning process
- incremental learning
- supervised learning
- pairwise
- bayesian networks
- model selection
- professional development
- cross validation
- em algorithm
- expectation maximization
- maximum likelihood
- probabilistic model
- support vector
- probability density function
- least squares
- prior knowledge
- classroom practice