Cross-Utterance Conditioned VAE for Non-Autoregressive Text-to-Speech.
Yang LiCheng YuGuangzhi SunHua JiangFanglei SunWeiqin ZuYing WenYang YangJun WangPublished in: CoRR (2022)
Keyphrases
- autoregressive
- text to speech
- non stationary
- speech synthesis
- moving average
- gaussian markov random field
- text to speech synthesis
- random fields
- programming tool
- speech recognition
- prosodic features
- sar images
- autoregressive model
- word processing
- random field models
- spoken language
- spectrum analysis
- writing skills
- prior knowledge
- arma model
- parameter estimation