AI Paper Atlas

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

From Structural Feedback to Prompt Policies: Learning Faithful Text-to-Image Prompt Editors

Wentao Ye, Yali Ye, Zhiqing Xiao, Ru Peng, Xinpeng Ti, Liyao Li, Zhanming Shen, Peng Lu, Jie Zhang, Haobo Wang

NeurIPS 2026 · Image generation › Text-to-image diffusion · approach: T2I evaluation and controllable adapters · 0 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted summary. Written by AI from the paper's full text. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Proposes learning faithful text-to-image prompt editors that edit user prompts guided by structural feedback rather than open-ended rewriting that may drift from user intent.

Problem

Text-to-image prompt optimization by open-ended rewriting can silently weaken the user's intent, especially for prompts specifying fine-grained objects, attributes, and relations.

Approach

Reframes prompt optimization as faithful prompt editing that learns a prompt policy from structural feedback on whether the user's specified objects, attributes, and relations are preserved in generated images.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add DSE scene-graph calibration (ablation)StruPO · LexicaSGR (%): 73.69 -> 80.74Improved
add graph encoder (MHE) (ablation)StruPO · LexicaSGR (%): 75.33 -> 80.74Improved
use edit actions (delete/reorder/insert) instead of free-form generation (ablation)StruPO · LexicaSGR (%): 73.68 -> 80.74Improved
add image-grounded RL after SFT (ablation)StruPO · LexicaSGR (%): 74.04 -> 80.74Improved
add structured consistency reward r_con (ablation)StruPO · LexicaSGR (%): 75.14 -> 80.74Improved
add structured consistency reward r_con (ablation)StruPO · LexicaAES_delta: .68 -> .63Hurt
add RL after SFT (ablation)StruPO · LexicaAES_delta: .57 -> .63Improved
add DSE calibration (ablation)StruPO · LexicaAES_delta: .64 -> .63No clear effect
StruPO vs best baseline on relation recallSD v1.5 · LexicaRR (%): 7.96 points over best listed baselineImproved
apply StruPO editor trained on SD v1.5 to unseen generators without retrainingStruPO · Lexica (SD v1.4, SD v2.1, FLUX.1-dev)RR / SGR: FLUX.1-dev RR 62.43 vs Promptist 53.41Improved
StruPO versus VisualPrompter on out-of-domain promptsStruPO · COCOSGR (%): 78.94 vs 83.26Hurt
StruPO versus VisualPrompter on out-of-domain promptsStruPO · COCOAES_delta: 1.02 vs -0.36Improved

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