AI Paper Atlas

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

Context Diffusion: In-Context Aware Image Generation

Ivona Najdenkoska, Animesh Sinha, Abhimanyu Dubey, Dhruv Mahajan, Vignesh Ramanathan, Filip Radenovic

ECCV 2024 · Image generation › Text-to-image diffusion · approach: Compositional text-to-image guidance · 14 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
inject visual context via cross-attention alongside text instead of Prompt Diffusion-style conditioningContext Diffusion · Human eval in-domain, context onlywin rate (%): 4.5 -> 80.2Improved
inject visual context via cross-attention vs Prompt Diffusion on out-of-domain tasksContext Diffusion · Human eval out-of-domain, context+promptwin rate (%): 26.9 -> 52.3Improved
prompt-only conditioning on in-domain tasksContext Diffusion · Human eval in-domain, prompt onlywin rate (%): 30.4 -> 29.6No clear effect
prompt-only conditioning on out-of-domain tasksContext Diffusion · Human eval out-of-domain, prompt onlywin rate (%): 25.9 -> 49.8Improved
use three context images instead of one (ablation)Context Diffusion · Human eval out-of-domain few-shot, context onlywin rate (%): 28.3 -> 50.2Improved
use three context images instead of one with prompt (ablation)Context Diffusion · Human eval out-of-domain few-shot, context+promptwin rate (%): 21.5 -> 60.0Improved
pair source image with target as context (as in Prompt Diffusion) (ablation)Context Diffusion · qualitativeimage qualityNo clear effect
encode context with ConvNets summed to the query image instead of visual conditioning in cross-attention (ablation)Context Diffusion · qualitativecontext fidelityHurt
context-only conditioning for depth-to-imageContext Diffusion · depth-to-image, context onlyFID: 27.0 -> 18.5Improved
prompt-only conditioning for seg-to-imageContext Diffusion · seg-to-image, prompt onlyFID: 17.2 -> 18.5Hurt

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