AI Paper Atlas

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

Aligning Multi-Character Narrative Image Generation with Multi-Aspect Human Preferences

Ziyi Gao, Zhipeng Wei, Jingjing Chen, Zhiyu Tan, Hao Li, Yi-Ping Phoebe Chen

CVPR 2026 · Image generation › Text-to-image diffusion · approach: T2I evaluation and controllable adapters

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 critique generation before scoring in the reward model (ablation)NIReward · NI-Bench visual qualitypreference accuracy (%): 72.66 -> 83.00Improved
add critique generation before scoring (ablation)NIReward · NI-Bench prompt followingpreference accuracy (%): 83.83 -> 86.07Improved
add critique generation before scoring (ablation)NIReward · NI-Bench identity consistencypreference accuracy (%): 84.60 -> 85.10Improved
train NIReward instead of using GPT-4o-mini as judgeNIReward · NI-Bench identity consistencypreference accuracy (%): 46.75 -> 85.10Improved
use NIReward instead of GPT-4o-mini on visual qualityNIReward · NI-Bench visual qualitypreference accuracy (%): 88.42 -> 83.00Hurt
add rejection sampling to ADPO (ablation)ADPO · Preference optimizationImageReward / NIReward P.F.: w/o RS 0.951 -> 1.010 ImageReward; 0.001 -> 0.079 P.F.Improved
add adaptive weighted learning to ADPO (ablation)ADPO · Preference optimizationNIReward identity consistency: -0.013 -> 0.010Improved
remove adaptive weighting (w/o AW) (ablation)ADPO · Preference optimizationNIReward prompt following: 0.079 (full) vs 0.123 (w/o AW)Mixed
use Dominating Comparison instead of mean scores (ablation)ADPO · Preference optimizationID Sim: 26.94 -> 28.88Improved
use ADPO with NIReward instead of Diffusion-DPO baselinesPhotoMaker V2 · Preference optimizationNIReward identity consistency: -0.036 -> 0.010Improved
optimize with ADPO with NIRewardPhotoMaker V2 · Preference optimizationCLIP: 32.38 -> 32.80Improved

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