Rapid Style Adaptation Using Residual Error Embedding for Expressive Speech Synthesis.
Xixin WuYuewen CaoMu WangSongxiang LiuShiyin KangZhiyong WuXunying LiuDan SuDong YuHelen MengPublished in: INTERSPEECH (2018)
Keyphrases
- speech synthesis
- residual error
- speech recognition
- speaker adaptation
- vocal tract
- text to speech
- prosodic features
- automatic speech recognition
- language model
- speech signal
- noisy environments
- speech corpus
- cost function
- hidden markov models
- vector space
- nonlinear dimensionality reduction
- neural network
- data hiding
- image acquisition
- probabilistic model
- image processing
- data mining