AI Paper Atlas

Research map / Image generation / Fast diffusion sampling

Ambient Diffusion: Learning Clean Distributions from Corrupted Data

Giannis Daras, Kulin Shah, Yuval Dagan, Aravind Gollakota, Alex Dimakis, Adam Klivans

NeurIPS 2023 · Image generation › Fast diffusion sampling · approach: Diffusion theory and design space · 10 citations (OpenAlex)

Read the paper

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add further corruption (extra masking) to training inputs and compute loss on all originally observed pixels (ablation)Ambient Diffusion · CIFAR-10 high corruptionFIDImproved
train on corrupted data with Ambient Diffusion vs AmbientGANCIFAR-10 generation · CIFAR-10 all corruption levelsgeneration quality (FID/Inception)Improved
single-step Ambient Diffusion restoration vs DPS (100 NFE) at p=0.6CelebA-HQ inpainting · CelebA-HQ random inpainting p=0.6PSNR: 28.21 -> 31.51Improved
single-step Ambient Diffusion restoration vs DPS (100 NFE) at p=0.4AFHQ inpainting · AFHQ random inpainting p=0.4PSNR: 34.06 -> 33.27Hurt
Ambient Diffusion vs supervised MAE inpaintingAFHQ inpainting · AFHQ random inpainting p=0.4PSNR: 28.88 -> 33.27Improved
add Reconstruction Guidance term to Fixed Mask sampler (ablation)Ambient Diffusion sampler · CIFAR-10 FIDFID: 11.70 -> 11.59 (p=0.2); 46.27 -> 46.31 (p=0.8)Mixed
fine-tune Deepfloyd IF on 80% pixel-deleted CelebA images vs standard fine-tuning on clean images (ablation)Deepfloyd IF fine-tuning · CelebA 3000-image subsetmemorization (DINOv2 nearest-neighbor similarity)Improved
increase corruption level in training data (ablation)Ambient Diffusion · CelebA full datasetmemorization vs sample qualityMixed

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy