Parameter Generation Methods With Rich Context Models for High-Quality and Flexible Text-To-Speech Synthesis.
Shinnosuke TakamichiTomoki TodaYoshinori ShigaSakriani SaktiGraham NeubigSatoshi NakamuraPublished in: IEEE J. Sel. Top. Signal Process. (2014)
Keyphrases
- high quality
- predictive power
- mathematical models
- statistical models
- statistical methods
- linear models
- ground truth
- linear regression
- benchmark datasets
- text to speech synthesis
- neural network
- qualitative and quantitative
- machine learning methods
- data mining techniques
- probabilistic model
- significant improvement
- high dimensional
- computer vision
- genetic algorithm