ML//diffusion model//DDPM
- Denoising Diffusion Probabilistic Models (Ho et al., 2020)
Denoising Diffusion Probabilistic Models (Ho et al., 2020)
The foundational paper that showed diffusion can match GANs on image quality.
U-Net architecture predicts the noise to subtract at each step.
Forward process adds Gaussian noise gradually, reverse process learns to undo it.