Semi-Supervised Acoustic Model Training for Five-Lingual Code-Switched ASR.
Astik BiswasEmre YilmazFebe de WetEwald van der WesthuizenThomas NieslerPublished in: INTERSPEECH (2019)
Keyphrases
- semi supervised
- supervised learning
- training process
- semi supervised learning
- fully labeled
- labeled data
- training set
- training samples
- data sets
- batch mode
- semi supervised classification
- co training
- active learning
- unsupervised learning
- unlabeled data
- machine learning
- source code
- supervised methods
- face recognition
- loss function
- multi view
- automatic speech recognition
- weakly supervised
- pairwise constraints
- prior knowledge
- probabilistic model