Improving Robustness of Speaker Recognition to New Conditions Using Unlabeled Data.
Diego CastánMitchell McLarenLuciana FerrerAaron LawsonAlicia Lozano-DiezPublished in: INTERSPEECH (2017)
Keyphrases
- unlabeled data
- speaker recognition
- labeled data
- semi supervised learning
- semi supervised
- gaussian mixture model
- active learning
- co training
- supervised learning
- training data
- vector quantization
- data points
- text classification
- speaker verification
- probabilistic neural network
- labeled and unlabeled data
- training set
- speaker identification
- learning algorithm
- prior knowledge
- speech recognition
- speech signal
- transfer learning
- decision trees
- clustering algorithm
- information extraction
- face recognition
- probabilistic model
- machine learning
- principal component analysis