AI Paper Atlas

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

Synthetic Curriculum Reinforces Compositional Text-to-Image Generation

Shijian Wang, Runhao Fu, Siyi Zhao, Qingqin Zhan, Xingjian Wang, Jiarui Jin, Yuan Lu, Hanqian Wu, Cunjian Chen

CVPR 2026 · Image generation › Text-to-image diffusion · approach: Layout-to-image control · 1 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
apply CompGen curriculum RL to Stable-Diffusion-1.5Stable-Diffusion-1.5 · 5 benchmarks avgavg score: 54.90% -> 66.62%Improved
apply CompGen curriculum RL to SimpleARSimpleAR-0.5B · 5 benchmarks avgavg score: 63.66% -> 71.27%Improved
use stronger MLLM reward model (LLaVA-v1.6-13B vs InstructBLIP) (ablation)CompGen reward · SD-1.5 5 benchmarks avgavg score: 57.22% -> 66.62%Improved
use CompGen fine-grained scene-graph questions with curriculum instead of plain VQAScore reward (ablation)SD-1.5 RL · 5 benchmarks avgavg score: 61.44% -> 66.62%Improved
use multiplicative scene-graph difficulty instead of additive (ablation)CompGen difficulty measure · SD-1.5, 10K promptsavg score: 62.06% (approx) -> 66.62%; paper states +4.56 pointsImproved
use multiplicative difficulty instead of additive on T2I-CompBench (ablation)CompGen difficulty measure · T2I-CompBenchscore: 37.00% -> 37.68%No clear effect
add random curriculum sampling vs no curriculum (ablation)CompGen GRPO · GenEval SD-1.5accuracy: 42% -> 52.1%Improved
use Gaussian scheduling instead of random sampling (ablation)CompGen GRPO · GenEval SD-1.5accuracy: 52.1% -> 54.6%Improved
train beyond ~500 steps (ablation)CompGen GRPO · GenEval SD-1.5accuracyHurt
train only on skew-easy or skew-hard difficulty distributions (ablation)CompGen · difficulty distributioncompositional generalizationHurt

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