AI Paper Atlas

Research map / Image generation / Fast diffusion sampling

Retrieval-Augmented Diffusion Models

Andreas Blattmann, Robin Rombach, Kaan Oktay, Jonas Müller, Björn Ommer

NeurIPS 2022 · Image generation › Fast diffusion sampling · approach: Diffusion models as perception tools · 32 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
train with large diverse OpenImages database instead of small or different-domain database (ablation)RDM on ImageNet dogs subset · ImageNet dogsFID/precision/recallImproved
use database from different domain (WikiArt) during training (ablation)RDM on ImageNet dogs subset · ImageNet dogssample qualityHurt
train database disjoint from train set (OpenImages) instead of train set itself, then swap inference database (ablation)RDM · ImageNet train CLIP-FIDCLIP-FID: 8.86 -> 3.84Improved
disjoint train database (OpenImages) vs train set as database (ablation)RDM · COCO zero-shot text-to-imageFID: 27.28 -> 22.08Improved
increase k_train from 1 to 2-8 neighbors (ablation)RDM · ImageNetFID / ISNo clear effect
increase k_train beyond 8 (to 16) (ablation)RDM · ImageNetrecallHurt
increase k_train up to 8 (ablation)RDM · COCO zero-shot text-to-imageFID / CLIP scoreImproved
add classifier-free guidance (s=1.75, m=0.1) to RDM-OI (ablation)RDM-OI · ImageNet unconditional train FIDFID: 24.50 -> 13.22Improved
top-m sampling truncation (smaller m) (ablation)RDM-OI · ImageNet unconditionalFID: 24.50 -> 19.08Improved
replace nearest-neighbor retrieval with direct CLIP image embedding conditioning plus flow prior (ablation)RDM inversion model · COCO zero-shot text-to-imageFID: ~87 -> ~45Improved
replace parametric LDM with retrieval-augmented RDM-OILDM · FFHQCLIP-FID: 2.12 -> 1.92Mixed
use more complex ImageNet subsets with semi-parametric modelsRDM and RARM · ImageNet subsetsFID/recallImproved

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