Using speaker adaptive training to realize Mandarin-Tibetan cross-lingual speech synthesis.
Hongwu YangKeiichiro OuraHaiyan WangZhenye GanKeiichi TokudaPublished in: Multim. Tools Appl. (2015)
Keyphrases
- speech synthesis
- cross lingual
- prosodic features
- speech recognition
- text to speech
- language modeling
- language model
- vocal tract
- machine translation
- cross lingual information retrieval
- event extraction
- cross language
- automatic speech recognition
- translation model
- speaker verification
- text classification
- parallel corpus
- news articles
- speaker identification
- parallel corpora
- hidden markov models
- pattern recognition
- information retrieval
- neural network
- statistical machine translation
- document clustering
- training set
- training data
- retrieval model