Can fitting the denoising data distribution of DDIM using conditional GAN networks in DDGANs further improve the generation speed？

NVlabs / denoising-diffusion-gan

Tackling the Generative Learning Trilemma with Denoising Diffusion GANs https://arxiv.org/abs/2112.07804

Other

698 stars 77 forks source link

Can fitting the denoising data distribution of DDIM using conditional GAN networks in DDGANs further improve the generation speed？ #43

Open uuuaziQAQ opened 8 months ago

uuuaziQAQ commented 8 months ago

Hello! I have a question and I hope to discuss it with you DDGANs abandon the assumption that the denoising distribution is Gaussian and use a conditional GAN to simulate this denoising distribution. So, the acceleration model of DDPM (which actually only modified the sampling algorithm), such as DDIM, also has a data distribution and a non Markov chain for denoising. Can the conditional GANs in DDGANs fit the denoising distribution of DDIM, and will this further improve the generation speed