Research map / Image generation / Text-to-image diffusion
PosterReward: Unlocking Accurate Evaluation for High-Quality Graphic Design Generation
Jianyu Lai, Sixiang Chen, Jialin Gao, Hengyu Shi, Zhongying Liu, Fuxiang Zhai, Junfeng Luo, Xiaoming Wei, Lujia Wang, Lei Zhu
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 changed | Where | Measured | Effect |
|---|---|---|---|
| add analysis module before the scoring head (ablation) | PosterReward-Lite · PRB-Basic | accuracy: 83.9 -> 85.7 | Improved |
| add analysis module before the scoring head (ablation) | PosterReward-Lite · PRB-Advanced | accuracy: 85.0 -> 85.8 | Improved |
| add GRPO training of the analysis module (ablation) | PosterReward (with analysis) · PRB-Basic | accuracy: 85.7 -> 86.7 | Improved |
| add GRPO training of the analysis module (ablation) | PosterReward (with analysis) · PRB-Advanced | accuracy: 85.8 -> 86.0 | No clear effect |
| add analysis module and GRPO (ablation) | PosterReward-Lite · HPDv3 | accuracy: 77.1 -> 77.8 | Improved |
| use joint (analysis + comparison) SFT instead of single-task SFT (ablation) | PosterReward-Pairwise · PRB-Advanced | average accuracy: 81.93 -> 82.71 | Improved |
| add joint rejection-sampling fine-tuning after joint SFT (ablation) | PosterReward-Pairwise · PRB-Advanced | average accuracy: 82.71 -> 83.82 | Improved |
| add joint rejection-sampling fine-tuning after joint SFT (ablation) | PosterReward-Pairwise · PRB-Basic | average accuracy: 81.92 -> 82.98 | Improved |
| use PosterReward instead of HPSv3 | HPSv3 · PRB-Advanced | accuracy: 41.2 -> 86.0 | Improved |
| use PosterReward instead of PosterReward-Lite | PosterReward-Lite · MMRB2 | accuracy: 60.5 -> 59.6 | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.
- HPSv3 — HPSv3: Towards Wide-Spectrum Human Preference Score (ICCV 2025) · better on 4 of 4 shared results
- ImageReward — ImageReward: Learning and Evaluating Human Preferences for Text-to-Image Generation (NeurIPS 2023) · better on 1 of 1 shared result
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
- P2P: Automated Paper-to-Poster Generation and Fine-Grained Benchmark (ICLR 2026 · similarity 0.89)
- POSTA: A Go-to Framework for Customized Artistic Poster Generation (CVPR 2025 · similarity 0.88)
- Can Vision–Language Models Assess Graphic Design Aesthetics? A Benchmark, Evaluation, and Dataset Perspective. (ICLR 2026 · similarity 0.88)
- Paper2Poster: Towards Multimodal Poster Automation from Scientific Papers (NeurIPS 2025 · similarity 0.87)
- AesthetiQ: Enhancing Graphic Layout Design via Aesthetic-Aware Preference Alignment of Multi-modal Large Language Models (CVPR 2025 · similarity 0.86)
- GlyphDraw2: Automatic Generation of Complex Glyph Posters with Diffusion Models and Large Language Models (AAAI 2025 · similarity 0.86)
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.
