Abstract: Recently, GAN vocoders have seen rapid progress in speech synthesis, starting to outperform autoregressive models in perceptual quality with much higher generation speed. However, ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results
Feedback