Research map / Image generation / Text-to-image diffusion
AutoPP: Towards Automated Product Poster Generation and Optimization
Jiahao Fan, Yuxin Qin, Wei Feng, Yanyin Chen, Yaoyu Li, Ao Ma, Yixiu Li, Li Zhuang, Haoyi Bian, Zheng Zhang, Jingjing Lv, Junjie Shen and 1 more
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 |
|---|---|---|---|
| use isolated DPO (element-weighted likelihood) instead of standard DPO (ablation) | unified design module · online CTR on JD.COM | Relative CTR: 3.10% -> 4.49% | Improved |
| replace decomposed attention with full attention (ablation) | element rendering module · visual text generation | Sentence accuracy: 65.19 vs 66.37 | No clear effect |
| use decomposed attention instead of full attention (ablation) | MM-DiT blocks · 800x800 / 1024x1024 resolution | GFLOPs: -18% (800x800), -24% (1024x1024) | Improved |
| increase optimizer training data from 10K to 30K to 50K (ablation) | IDPO optimizer · reward accuracy, 1,000 test samples | Reward accuracy: 51.20% -> 67.19% -> 75.99% | Improved |
| encode glyph image as tokens instead of pixel-aligned ControlNet or inpainting control | element rendering module · visual text generation | Sentence accuracy: 51.20 -> 65.19 (vs FLUX-ControlNet) | Improved |
| use a unified MLLM design module instead of GPT-4o with same product info | poster generation · AutoPP1M poster generation | FID: 63.47 -> 60.71 | Improved |
| CTR optimization of backgrounds only (CG4CTR, CAIG) without layout and text | poster generator · online CTR on JD.COM | Relative CTR | Hurt |
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.
- PosterMaker — PosterMaker: Towards High-Quality Product Poster Generation with Accurate Text Rendering (CVPR 2025) · better on 2 of 2 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
- PosterMaker: Towards High-Quality Product Poster Generation with Accurate Text Rendering (CVPR 2025 · similarity 0.87)
- P2P: Automated Paper-to-Poster Generation and Fine-Grained Benchmark (ICLR 2026 · similarity 0.86)
- POSTA: A Go-to Framework for Customized Artistic Poster Generation (CVPR 2025 · similarity 0.85)
- Paper2Poster: Towards Multimodal Poster Automation from Scientific Papers (NeurIPS 2025 · similarity 0.84)
- GlyphDraw2: Automatic Generation of Complex Glyph Posters with Diffusion Models and Large Language Models (AAAI 2025 · similarity 0.84)
- Desigen: A Pipeline for Controllable Design Template Generation (CVPR 2024 · similarity 0.84)
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.
