AI Paper Atlas

Research map / Image generation / Text-to-image diffusion

Zero-Shot Spatial Layout Conditioning for Text-to-Image Diffusion Models

Guillaume Couairon, Marlène Careil, Matthieu Cord, Stéphane Lathuilière, Jakob Verbeek

ICCV 2023 · Image generation › Text-to-image diffusion · approach: Compositional text-to-image guidance · 68 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

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 ZestGuide attention-based guidance over PwWLDM · COCO-Stuff Eval-fewmIoU: 36.3 -> 46.9Improved
use ZestGuide instead of external-classifier guidanceLDM · COCO-Stuff Eval-fewmIoU / time: mIoU 20.5 -> 46.9; 1 min -> around 15 secs per imageImproved
train-based methods evaluated with only a few segmentsOASIS, SDM · COCO-Stuff Eval-fewmIoU: OASIS 52.1 (Eval-all) -> 41.4 (Eval-few)Hurt
average attention across all heads before applying the loss (ablation)ZestGuide · COCO-Stuff Eval-fewmIoU: 32.1 -> 43.3Improved
average attention across layers instead of per-layer loss (ablation)ZestGuide · COCO-Stuff Eval-fewmIoU: 42.7 -> 43.3Improved
add L-infinity gradient normalization (ablation)ZestGuide · COCO-Stuff Eval-fewmIoU: 38.99 -> 43.34Improved
apply guidance for all steps (tau=1) instead of first half (ablation)ZestGuide · COCO-Stuff Eval-fewmIoU / FID: mIoU 43.34 -> 44.58, FID 31.53 -> 34.75Mixed
reduce guidance to first 10% of steps (tau=0.1) (ablation)ZestGuide · COCO-Stuff Eval-fewmIoU: 43.34 -> 34.25Hurt
restrict attention layers to res16 and res32 (ablation)ZestGuide · COCO-Stuff Eval-fewmIoU: 40.17 -> 43.34Improved
combine the normalized-map loss with PwWZestGuide · COCO-Stuff Eval-filteredFID-mIoU trade-offImproved
use class-text embeddings as keys instead of full prompt attention (ablation)ZestGuide · COCO-Stuff Eval-fewmIoU / FID: mIoU 43.36 vs 43.34, FID about +3Hurt

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 0 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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