A Fully Time-domain Neural Model for Subband-based Speech Synthesizer.
Azam RabieeSoo-Young LeePublished in: CoRR (2018)
Keyphrases
- subband
- neural model
- frequency domain
- text to speech
- filter bank
- linear prediction
- wavelet transform
- neural network
- image compression
- multiresolution
- recurrent neural networks
- feature vectors
- bit rate
- wavelet coefficients
- control scheme
- wavelet decomposition
- discrete wavelet transform
- spatial domain
- high frequency
- low pass
- speech recognition
- subband coding
- subband decomposition
- image coding
- multi layer perceptron
- image coder
- wavelet coding
- perfect reconstruction
- speech signal
- wavelet packet
- rate distortion
- multiscale
- pattern recognition
- dynamic model
- video signals
- real time
- image processing
- wavelet subbands
- closed loop
- denoising