Prediction of Creaky Speech by Recurrent Neural Networks Using Psychoacoustic Roughness.
Julián VillegasKonstantin MarkovJeremy PerkinsSeunghun J. LeePublished in: IEEE J. Sel. Top. Signal Process. (2020)
Keyphrases
- recurrent neural networks
- chaotic time series
- neural network
- prediction accuracy
- feed forward
- audio signals
- echo state networks
- speech recognition
- reservoir computing
- complex valued
- feedforward neural networks
- neural model
- cascade correlation
- artificial neural networks
- recurrent networks
- linear prediction
- prediction model
- prediction error
- long short term memory
- text to speech
- speech signal
- automatic speech recognition
- nonlinear dynamic systems