Leveraging Pseudo-labeled Data to Improve Direct Speech-to-Speech Translation.
Qianqian DongFengpeng YueTom KoMingxuan WangQibing BaiYu ZhangPublished in: INTERSPEECH (2022)
Keyphrases
- labeled data
- semi supervised learning
- semi supervised
- unlabeled data
- training data
- active learning
- text classification
- speech recognition
- domain adaptation
- semi supervised classification
- supervised learning
- labeled examples
- transfer learning
- label propagation
- prior knowledge
- co training
- labeled training data
- data points
- class labels
- neural network
- labeled and unlabeled data
- labeled instances
- target domain
- training examples
- natural language
- learning algorithm
- machine learning