Cross-Lingual Speaker Adaptation Using Domain Adaptation and Speaker Consistency Loss for Text-To-Speech Synthesis.
Detai XinYuki SaitoShinnosuke TakamichiTomoki KoriyamaHiroshi SaruwatariPublished in: Interspeech (2021)
Keyphrases
- cross lingual
- domain adaptation
- speaker adaptation
- sentiment classification
- speech recognition
- transfer learning
- maximum likelihood
- language modeling
- automatic speech recognition
- language model
- labeled data
- machine translation
- text classification
- cross domain
- parallel corpora
- speaker independent
- training data
- semi supervised
- text to speech
- speech signal
- unlabeled data
- news articles
- neural network
- pattern recognition
- semi supervised learning
- speaker identification
- document classification
- feature space
- pairwise
- broadcast news
- collaborative filtering
- natural language processing
- active learning
- face recognition