Initial investigation of speech synthesis based on complex-valued neural networks.
Qiong HuJunichi YamagishiKorin RichmondKartick SubramanianYannis StylianouPublished in: ICASSP (2016)
Keyphrases
- complex valued
- speech synthesis
- neural network
- activation function
- recurrent neural networks
- real valued
- speech recognition
- text to speech
- artificial neural networks
- vocal tract
- pattern recognition
- fuzzy logic
- back propagation
- singular spectrum analysis
- feed forward
- hidden layer
- neural nets
- network architecture
- multi valued
- multi layer perceptron
- linear equations
- multilayer perceptron
- neuro fuzzy
- neural network model
- radial basis function
- random variables
- supervised learning
- genetic algorithm