Application-dependent prosodic models for text-to-speech synthesis and automatic design of learning database corpus using genetic algorithm.
Olivier BoëffardFrançoise EmerardPublished in: EUROSPEECH (1997)
Keyphrases
- text to speech synthesis
- database
- application dependent
- genetic algorithm
- knowledge acquisition
- prior knowledge
- databases
- learning process
- learning algorithm
- database systems
- probabilistic model
- learning systems
- relational databases
- neural network
- underlying assumptions
- accurate models
- structured prediction
- hidden variables
- learning design
- learning models
- design principles
- supervised learning
- active learning
- fitness function
- semi automatic
- metadata
- online learning
- object oriented
- hybrid learning
- multi objective
- learned models
- case study
- artificial neural networks
- reinforcement learning