---
license: cc-by-nc-sa-4.0
tags:
  - sketch-colorization
  - diffusion
  - stable-diffusion-xl
  - image-to-image
  - reference-based
---

# ColorizeDiffusion XL - Model Weights

<div align="center">

[![WACV 2025](https://img.shields.io/badge/WACV%202025-v1-0CA4A5?style=flat&logo=Semantic%20Web)](https://openaccess.thecvf.com/content/WACV2025/html/Yan_ColorizeDiffusion_Improving_Reference-Based_Sketch_Colorization_with_Latent_Diffusion_Model_WACV_2025_paper.html)
[![CVPR 2025](https://img.shields.io/badge/CVPR%202025-v1.5-0CA4A5?style=flat&logo=Semantic%20Web)](https://arxiv.org/abs/2502.19937)
[![CVPR 2026](https://img.shields.io/badge/CVPR%202026-XL-0CA4A5?style=flat&logo=Semantic%20Web)](https://arxiv.org/abs/2603.05971)
[![GitHub](https://img.shields.io/badge/GitHub-Code-181717?style=flat&logo=GitHub)](https://github.com/tellurion-kanata/ColorizeDiffusionXL)
[![Demo](https://img.shields.io/badge/Hugging%20Face-Demo-FF9D00?style=flat&logo=Hugging%20Face)](https://huggingface.co/spaces/tellurion/ColorizeDiffusion)

</div>

Model weights for [ColorizeDiffusion XL](https://github.com/tellurion-kanata/ColorizeDiffusionXL), an SDXL-based reference-based sketch colorization framework (1024px).
For the base SD2.1 implementation (512/768px), refer to the [original repository](https://github.com/tellurion-kanata/colorizeDiffusion) and [weights](https://huggingface.co/tellurion/ColorizeDiffusion).

## Model Variants

| Filename prefix | Config | Description |
|:----------------|:-------|:------------|
| `sdxl-*` | `configs/inference/sdxl.yaml` | Base SDXL colorizer with embedding guidance |
| `xlv2-*` | `configs/inference/xlv2.yaml` | SDXL colorizer with ReferenceNet adapter for enhanced background/style |

## Usage

### Inference with UI

```bash
git clone https://github.com/tellurion-kanata/ColorizeDiffusionXL.git
cd ColorizeDiffusionXL
conda env create -f environment.yml
conda activate hf
python -u app.py
```

Select and load a checkpoint from the UI dropdown. Weights are downloaded automatically from this repository.

### Online Demo

Try the model directly on [Hugging Face Spaces](https://huggingface.co/spaces/tellurion/ColorizeDiffusion).

### Programmatic Download

```python
from huggingface_hub import hf_hub_download

path = hf_hub_download(
    repo_id="tellurion/ColorizeDiffusionXL",
    filename="<checkpoint_name>.safetensors",
)
```

## License

These weights are released under [CC-BY-NC-SA 4.0](https://creativecommons.org/licenses/by-nc-sa/4.0/).

## Citation

```bibtex
@InProceedings{Yan_2025_WACV,
    author    = {Yan, Dingkun and Yuan, Liang and Wu, Erwin and Nishioka, Yuma and Fujishiro, Issei and Saito, Suguru},
    title     = {ColorizeDiffusion: Improving Reference-Based Sketch Colorization with Latent Diffusion Model},
    booktitle = {Proceedings of the Winter Conference on Applications of Computer Vision (WACV)},
    year      = {2025},
    pages     = {5092-5102}
}

@article{2025arXiv250219937Y,
    author = {{Yan}, Dingkun and {Wang}, Xinrui and {Li}, Zhuoru and {Saito}, Suguru and {Iwasawa}, Yusuke and {Matsuo}, Yutaka and {Guo}, Jiaxian},
    title = "{Image Referenced Sketch Colorization Based on Animation Creation Workflow}",
    journal = {arXiv e-prints},
    year = {2025},
    doi = {10.48550/arXiv.2502.19937},
}

@article{2026arXiv260305971Y,
    author = {{Yan}, Dingkun and {Wang}, Xinrui and {Iwasawa}, Yusuke and {Matsuo}, Yutaka and {Saito}, Suguru and {Guo}, Jiaxian},
    title = "{ColorizeDiffusion XL: Enhancing Embedding Guidance for Character Colorization and Geometry Disentanglement}",
    journal = {arXiv e-prints},
    year = {2026},
    doi = {10.48550/arXiv.2603.05971},
}
```