chrisdonahue / wavegan

WaveGAN: Learn to synthesize raw audio with generative adversarial networks
MIT License
1.33k stars 282 forks source link

How to convert a GAN to audio from image #82

Closed Tylersuard closed 4 years ago

Tylersuard commented 4 years ago

Hello. I'm reading your paper right now. It seems you converted DCGAN to run on a 1x16384 tensor rather than the usual image. I was wondering, how would one go about doing that, code-wise? Is it as simple as just changing the input dimensions? I tried to figure it out myself, but your code is very different from DCGAN. Looking forward to your response. Thank you, Tyler Suard