flâneur — a map of the web's best reading

[2006.11239] Denoising Diffusion Probabilistic Models

arxiv.org · 8,411 words · saved by 1 readers

Abstract: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 this https URL

# link_rsil0qba7t.pdf ## Metadata - PDFFormatVersion=1.5 - IsLinearized=false - IsAcroFormPresent=false - IsXFAPresent=false - IsCollectionPresent=false - IsSignaturesPresent=false - CreationDate=D:20201218013012Z - Creator=LaTeX with hyperref - ModDate=D:20201218013012Z - Custom.PTEX.Fullbanner=This is pdfTeX, Version 3.14159265-2.6-1.40.21 (TeX Live 2020) kpathsea version 6.3.2 - Producer=pdfTeX-1.40.21 - Trapped=False ## Contents ### Page 1 Denoising Diffusion Probabilistic ModelsJonathan HoUC Berkeleyjonathanho@berkeley.eduAjay JainUC Berkeleyajayj@berkeley.eduPieter AbbeelUC Berkeleypa

Explore this link on the map →

saved by

related reading