A Multi-Scale Time-Frequency Spectrogram Discriminator for GAN-based Non-Autoregressive TTS.
Haohan GuoHui LuXixin WuHelen MengPublished in: INTERSPEECH (2022)
Keyphrases
- autoregressive
- multiscale
- short time fourier transform
- wavelet transform
- text to speech
- non stationary
- moving average
- gaussian markov random field
- image segmentation
- fast fourier transform
- scale space
- random fields
- edge detection
- sar images
- spectral analysis
- frequency domain
- natural images
- image processing
- signal processing
- speech signal
- wavelet decomposition
- autoregressive model
- structuring elements
- multiresolution
- wavelet domain
- autoregressive moving average
- random field models
- image compression
- arma model
- least squares
- wavelet packet
- wavelet coefficients
- computer vision