Transfer Learning from Speech Synthesis to Voice Conversion with Non-Parallel Training Data.
Mingyang ZhangYi ZhouLi ZhaoHaizhou LiPublished in: CoRR (2020)
Keyphrases
- transfer learning
- speech synthesis
- training data
- text to speech
- labeled data
- speech recognition
- target domain
- domain adaptation
- prosodic features
- learning algorithm
- semi supervised learning
- learning tasks
- training and test data
- knowledge transfer
- active learning
- vocal tract
- cross domain
- machine learning
- decision trees
- source domain
- supervised learning
- unlabeled data
- reinforcement learning
- text classification
- machine learning algorithms
- multi task
- manifold alignment
- semi supervised
- test data
- data sets
- training examples
- training set
- collaborative filtering
- label information
- labeled training data
- transferring knowledge
- class labels
- transfer knowledge
- test set
- text mining
- domain specific
- previously learned
- prior knowledge
- unsupervised learning
- training samples
- pairwise
- text categorization
- pattern recognition