Denoising Diffusion Probabilistic Models

06/19/2020 ∙ by Jonathan Ho, et al. ∙ 7

We present high quality image synthesis results using diffusion probabilistic models, a class of latent variable models inspired by considerations from nonequilibrium thermodynamics. Our best results are obtained by training on a weighted variational bound designed according to a novel connection between diffusion probabilistic models and denoising score matching with Langevin dynamics, and our models naturally admit a progressive lossy decompression scheme that can be interpreted as a generalization of autoregressive decoding. On the unconditional CIFAR10 dataset, we obtain an Inception score of 9.46 and a state-of-the-art FID score of 3.17. On 256x256 LSUN, we obtain sample quality similar to ProgressiveGAN. Our implementation is available at https://github.com/hojonathanho/diffusion

READ FULL TEXT
POST COMMENT

Comments

There are no comments yet.

Authors

page 16

page 17

page 18

page 19

page 21

page 22

page 23

page 24

Code Repositories

diffusion

Denoising Diffusion Probabilistic Models


view repo

denoising-diffusion-pytorch

Implementation of Denoising Diffusion Probabilistic Model in Pytorch


view repo

ddpm-proteins

A denoising diffusion probabilistic model (DDPM) tailored for conditional generation of protein distograms


view repo

diffusion_models

A series of tutorial notebooks on denoising diffusion probabilistic models in PyTorch


view repo

denoising_diffusion

Unofficial Implementation of "Denoising Diffusion Probabilistic Models" in PyTorch(Lightning)


view repo
This week in AI

Get the week's most popular data science and artificial intelligence research sent straight to your inbox every Saturday.