AI Paper Atlas

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

Consistent Text-to-Image Generation via Scene De-Contextualization

Song Tang, Peihao Gong, Kunyu LI, Kai Guo, Boyu Wang, Mao Ye, Jianwei Dr. Zhang, Xiatian Zhu

ICLR 2026 · Image generation › Text-to-image diffusion · approach: Personalized and region-controlled generation

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
use soft learned estimation of the shared subspace instead of a hard correlation matrix (ablation)SDeC · ConsiStory+DreamSim-F: 0.3351 -> 0.2589Improved
use absolute eigenvalue excursion rather than relative normalization (ablation)SDeC · ConsiStory+DreamSim-F: 0.3576 -> 0.2589Improved
use soft learned estimation (ablation)SDeC · ConsiStory+DreamSim-B: 0.4254 -> 0.3675Hurt
add SDeC prompt embedding editing to SDXLSDXL · ConsiStory+DreamSim-F: 0.2778 -> 0.2589Improved
add SDeC prompt embedding editing to SDXLSDXL · ConsiStory+DreamSim-B: 0.3861 -> 0.3675Hurt
combine SDeC with ConsiStoryConsiStory · ConsiStory+DreamSim-F: 0.2729 -> 0.2542Improved
combine SDeC with ConsiStoryConsiStory · ConsiStory+CLIP-I: 0.8604 -> 0.8744Improved
replace 1P1S prompt re-structuring (w/o IPCA) with SDeC embedding editing1Prompt1Story w/o IPCA · ConsiStory+DreamSim-F: 0.2682 -> 0.2589Improved
place scene prompt before ID prompt (reverse order) (ablation)SDeC · ConsiStory+DreamSim-F: 0.2589 -> 0.2670Hurt
place scene prompt before ID prompt (reverse order) (ablation)SDXL · ConsiStory+DreamSim-F: 0.2778 -> 0.2894Hurt
use SDeC instead of training-based methodsconsistent T2I · ConsiStory+ user studywin rate (%): 42.67% vs 20.83% (ConsiStory)Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 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