PosterCraft

<div align="center"> <h1>🎨 PosterCraft:<br/>Rethinking High-Quality Aesthetic Poster Generation in a Unified Framework (ICLR 2026)</h1>

<img src="images/logo/logo2.png" alt="PosterCraft Logo" width="1000"/>

🌐 Website | 🎯 Demo | 📄 Paper | 🤗 Models | 📚 Datasets | 🎥 Video | 🤗 HF Demo

</div>

---

News & Updates

• 🧩 [2025.06] Community user @AIFSH has successfully integrated PosterCraft into ComfyUI!
  You can check out the full workflow here: PosterCraft-ComfyUI Example
  Big thanks to the contributor — this will be helpful for many users! See Issue #6 for details.
• 📖 [2025.06] Our Chinese article providing a detailed introduction and technical walkthrough of PosterCraft is now available!
  Read it here: 中文解读｜高质量美学海报生成框架 PosterCraft
• 🔥 [2025.06] We have deployed a demo on Hugging Face Space, feel free to give it a try!
• 🚀 [2025.06] Our gradio demo and inference code are now available!
• 📊 [2025.06] We have released partial datasets and model weights on HuggingFace.

---

👥 Authors

Sixiang Chen^1,2*, Jianyu Lai¹*, Jialin Gao²*, Tian Ye¹, Haoyu Chen¹, Hengyu Shi², Shitong Shao¹, Yunlong Lin³, Song Fei¹, Zhaohu Xing¹, Yeying Jin⁴, Junfeng Luo², Xiaoming Wei², Lei Zhu^1,5†

¹The Hong Kong University of Science and Technology (Guangzhou)
²Meituan
³Xiamen University
⁴National University of Singapore
⁵The Hong Kong University of Science and Technology

*Equal Contribution, †Corresponding Author

---

🌟 What is PosterCraft?

<div align="center"> <img src="images/demo/demo2.png" alt="What is PosterCraft - Quick Prompt Demo" width="1000"/> <br> </div>

PosterCraft is a unified framework for high-quality aesthetic poster generation that excels in precise text rendering, seamless integration of abstract art, striking layouts, and stylistic harmony.

🚀 Quick Start

🔧 Installation

# Clone the repository
git clone https://github.com/ephemeral182/PosterCraft.git
cd PosterCraft

# Create conda environment
conda create -n postercraft python=3.11
conda activate postercraft

# Install dependencies
pip install -r requirements.txt

🚀 Quick Generation

Generate high-quality aesthetic posters from your prompt with BF16 precision:

python inference.py \
  --prompt "Urban Canvas Street Art Expo poster with bold graffiti-style lettering and dynamic colorful splashes" \
  --enable_recap \
  --num_inference_steps 28 \
  --guidance_scale 3.5 \
  --seed 42 \
  --pipeline_path "black-forest-labs/FLUX.1-dev" \
  --custom_transformer_path "PosterCraft/PosterCraft-v1_RL" \
  --qwen_model_path "Qwen/Qwen3-8B"

If you are running on a GPU with limited memory, you can use inference_offload.py to offload some components to the CPU:

python inference_offload.py \
  --prompt "Urban Canvas Street Art Expo poster with bold graffiti-style lettering and dynamic colorful splashes" \
  --enable_recap \
  --num_inference_steps 28 \
  --guidance_scale 3.5 \
  --seed 42 \
  --pipeline_path "black-forest-labs/FLUX.1-dev" \
  --custom_transformer_path "PosterCraft/PosterCraft-v1_RL" \
  --qwen_model_path "Qwen/Qwen3-8B"

💻 Gradio Web UI

We provide a Gradio web UI for PosterCraft.

python demo_gradio.py

📊 Performance Benchmarks

<div align="center">

📈 Quantitative Results

<table> <thead> <tr> <th>Method</th> <th>Text Recall ↑</th> <th>Text F-score ↑</th> <th>Text Accuracy ↑</th> </tr> </thead> <tbody> <tr> <td style="white-space: nowrap;">OpenCOLE (Open)</td> <td>0.082</td> <td>0.076</td> <td>0.061</td> </tr> <tr> <td style="white-space: nowrap;">Playground-v2.5 (Open)</td> <td>0.157</td> <td>0.146</td> <td>0.132</td> </tr> <tr> <td style="white-space: nowrap;">SD3.5 (Open)</td> <td>0.565</td> <td>0.542</td> <td>0.497</td> </tr> <tr> <td style="white-space: nowrap;">Flux1.dev (Open)</td> <td>0.723</td> <td>0.707</td> <td>0.667</td> </tr> <tr> <td style="white-space: nowrap;">Ideogram-v2 (Close)</td> <td>0.711</td> <td>0.685</td> <td>0.680</td> </tr> <tr> <td style="white-space: nowrap;">BAGEL (Open)</td> <td>0.543</td> <td>0.536</td> <td>0.463</td> </tr> <tr> <td style="white-space: nowrap;">Gemini2.0-Flash-Gen (Close)</td> <td>0.798</td> <td>0.786</td> <td>0.746</td> </tr> <tr> <td style="white-space: nowrap;"><b>PosterCraft (ours)</b></td> <td><b>0.787</b></td> <td><b>0.774</b></td> <td><b>0.735</b></td> </tr> </tbody> </table> <img src="images/user_study/hpc.png" alt="User Study Results" width="1000"/> </div>

---

🎭 Gallery & Examples

<div align="center">

🎨 PosterCraft Gallery

<table> <tr> <td align="center"><img src="images/gallery/gallery_demo1.png" width="250"><br><b>Adventure Travel</b></td> <td align="center"><img src="images/gallery/gallery_demo2.png" width="250"><br><b>Post-Apocalyptic</b></td> <td align="center"><img src="images/gallery/gallery_demo3.png" width="250"><br><b>Sci-Fi Drama</b></td> </tr> <tr> <td align="center"><img src="images/gallery/gallery_demo4.png" width="250"><br><b>Space Thriller</b></td> <td align="center"><img src="images/gallery/gallery_demo5.png" width="250"><br><b>Cultural Event</b></td> <td align="center"><img src="images/gallery/gallery_demo6.png" width="250"><br><b>Luxury Product</b></td> </tr> <tr> <td align="center"><img src="images/gallery/gallery_demo7.png" width="250"><br><b>Concert Show</b></td> <td align="center"><img src="images/gallery/gallery_demo8.png" width="250"><br><b>Children's Book</b></td> <td align="center"><img src="images/gallery/gallery_demo9.png" width="250"><br><b>Movie Poster</b></td> </tr> </table> </div>

---

🏗️ Model Architecture

<div align="center"> <img src="images/overview/framework_fig.png" alt="PosterCraft Framework Overview" width="1000"/> <br> <em><strong>A unified framework for high-quality aesthetic poster generation</strong></em> </div>

Our unified framework consists of four critical optimization stages in the training workflow:

🔤 Stage 1: Text Rendering Optimization

Addresses accurate text generation by precisely rendering diverse text on high-quality backgrounds, also ensuring faithful background representation and establishing foundational fidelity and robustness for poster generation.

🎨 Stage 2: High-quality Poster Fine-tuning

Shifts focus to overall poster style and text-background harmony using Region-aware Calibration. This fine-tuning stage preserves text accuracy while strengthening the artistic integrity of the aesthetic poster.

🎯 Stage 3: Aesthetic-Text RL

Employs Aesthetic-Text Preference Optimization to capture higher-order aesthetic trade-offs. This reinforcement learning stage prioritizes outputs that satisfy holistic aesthetic criteria and mitigates defects in font rendering.

🔄 Stage 4: Vision-Language Feedback

Introduces a Joint Vision-Language Conditioning mechanism. This iterative feedback combines visual information with targeted text suggestions for multi-modal corrections, progressively refining aesthetic content and background harmony.

---

💾 Model Zoo

We provide the weights for our core models, fine-tuned at different stages of the PosterCraft pipeline.

<div align="center"> <table> <tr> <th>Model</th> <th>Stage</th> <th>Description</th> <th>Download</th> </tr> <tr> <td>🎯 <b>PosterCraft-v1_RL</b></td> <td>Stage 3: Aesthetic-Text RL</td> <td>Optimized via Aesthetic-Text Preference Optimization for higher-order aesthetic trade-offs.</td> <td><a href="https://huggingface.co/PosterCraft/PosterCraft-v1_RL">🤗 HF</a></td> </tr> <tr> <td>🔄 <b>PosterCraft-v1_Reflect</b></td> <td>Stage 4: Vision-Language Feedback</td> <td>Iteratively refined using vision-language feedback for further harmony and content accuracy.</td>