Sihan Xu*, Yidong Huang*, Jiayi Pan, Ziqiao Ma, Joyce Chai
University of Michigan, Ann Arbor
University of California, Berkeley

This code was tested with python 3.9, Pytorch 2.2.1 using pretrained models through huggingface / diffusers. Specifically, we implemented our method over LCM and prompt2prompt.

We implemented our method with diffusers with a similar code structure to prompt2prompt but we modified the code to adapt to newest diffusors version. You can download the requirements using

pip install -r requirements.txt

Online demos

We provide online demo with Gradio app. You can play with the it or clone it to your own device!

Local Gradio demo

You can launch the provided Gradio demo locally with

python app_infedit.py

For further information about the input parameters, please refer to the Handbook.

@article{xu2023infedit,
  title={Inversion-Free Image Editing with Natural Language}, 
  author={Sihan Xu and Yidong Huang and Jiayi Pan and Ziqiao Ma and Joyce Chai},
  booktitle={Conference on Computer Vision and Pattern Recognition 2024},
  year={2024}
}

We thank the awesome research works Prompt-to-Prompt, LCM and Masactrl.