🖌️ UniCalli-Dev - Chinese Calligraphy Generator

A Unified Diffusion Framework for Column-Level Generation and Recognition of Chinese Calligraphy

Links

• 🌐 Project Page: https://envision-research.github.io/UniCalli/
• 📄 Paper: arXiv:2510.13745
• 💻 Code: GitHub
• 🎨 Demo: https://huggingface.co/spaces/TSXu/UniCalli_Dev
• 🤗 Model: TSXu/Unicalli_Pro

Features

• 1-7 Chinese Characters: Supports generating 1 to 7 Chinese characters in a column
• Historical Masters: 90+ calligraphers including 王羲之, 颜真卿, 赵佶/宋徽宗, etc.
• Multiple Font Styles: 楷 (Regular), 行 (Running), 草 (Cursive)
• Interactive Session: Generate multiple images in one GPU session
• 4-bit Quantization: Runtime quantization for efficient inference on limited GPU memory

Usage

1. Enter 1-7 Chinese characters
2. Select a calligrapher (or use synthetic style)
3. Choose a font style
4. Click "Start Generation"

Citation

If you find UniCalli useful in your research, please cite our paper:

@article{xu2025unicalli,
  title={UniCalli: A Unified Diffusion Framework for Column-Level
         Generation and Recognition of Chinese Calligraphy},
  author={Xu, Tianshuo and Wang, Kai and Chen, Zhifei and Wu, Leyi
          and Wen, Tianshui and Chao, Fei and Chen, Ying-Cong},
  journal={arXiv preprint arXiv:2510.13745},
  year={2025}
}