A Multi-Scale Time-Frequency Spectrogram Discriminator for GAN-based Non-Autoregressive TTS.
Haohan GuoHui LuXixin WuHelen MengPublished in: CoRR (2022)
Keyphrases
- autoregressive
- multiscale
- short time fourier transform
- wavelet transform
- text to speech
- non stationary
- gaussian markov random field
- moving average
- random fields
- autoregressive model
- image segmentation
- spectral analysis
- natural images
- scale space
- edge detection
- speech signal
- sar images
- random field models
- image processing
- multiresolution
- fast fourier transform
- frequency domain
- signal processing
- image compression
- wavelet coefficients
- denoising
- maximum entropy
- hidden markov models
- color images
- image analysis
- autoregressive moving average
- feature extraction