ID-Patch-SDXL

Official Gradio Demo for Our CVPR 2025 Paper

"ID-Patch: Robust ID Association for Group Photo Personalization"

[Project Page]  [Paper]  [Poster]  [Code]  [Model]

💡 How to Use This Demo?

  1. Upload ID images:

    • Upload one or more ID images for each person you want to generate.
      (The number of uploaded ID images should match the number of people in your pose reference image.)

  2. ID Order:

    • List the ID images separated by commas, following the left-to-right order of detected faces in the pose reference image.
      (ID index starts from 0!)

  3. Upload a pose reference image:

    • Choose an image that shows the desired pose(s) for the people you want to generate. (Tip: If the pose is too complicated, then the face detection and pose detection might fail.)

  4. Enter a text prompt:

    • Describe the scene you want to create.
      (Tip: Try to match the interactions described in your text with the uploaded pose reference.)

  5. [Optional] Adjust advanced settings:
    Fine-tune generation details if needed.

  6. Click "Generate":
    Your personalized image will be created. Enjoy!

We also offer example data that users can easily select and load by clicking the “Load Example” button for testing.
Example Selection
Number of Men
Number of Women
Select Pose Example
0 1
0 1

📜 Disclaimer and Licenses

The images used in this demo are sourced from consented subjects or generated by the models. These pictures are intended solely to show the capabilities of our research. If you have any concerns, please contact us, and we will promptly remove any inappropriate content.

The use of the released code, model, and demo must strictly adhere to the respective licenses. Our code is released under the Apache License 2.0, and our model is released under the CreativeML Open RAIL++-M License for academic research purposes only. Any manual or automatic downloading of the face models from InsightFace, the Juggernaut-X-v10 base model, etc., must follow their original licenses and be used only for academic research purposes.

This research aims to positively impact the field of Generative AI. Any usage of this method must be responsible and comply with local laws. The developers do not assume any responsibility for any potential misuse. We added the "AI Generated: ID-Patch" watermark for enhanced safety.

📖 Citation

If you find ID-Patch useful for your research or applications, please cite our paper:

@InProceedings{zhang2025idpatch,
    author    = {Zhang, Yimeng and Zhi, Tiancheng and Liu, Jing and Sang, Shen and Jiang, Liming and Yan, Qing and Liu, Sijia and Luo, Linjie},
    title     = {ID-Patch: Robust ID Association for Group Photo Personalization},
    booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
    month     = {June},
    year      = {2025}
}

We also appreciate it if you could give a star ⭐ to our Github repository. Thanks a lot!