✂️ Follow-Your-Shape: Shape-Aware Image Editing via Trajectory-Guided Region Control

Zeqian Long^2*, Mingzhe Zheng^1*, Kunyu Feng^1*, Xinhua Zhang, Hongyu Liu¹, Harry Yang¹, Linfeng Zhang³, Qifeng Chen¹, Yue Ma¹,

¹ HKUST, ² UIUC, ³ Shanghai Jiao Tong University

We propose Follow-Your-Shape, a training-free and mask-free framework that supports precise and controllable editing of object shapes while strictly preserving non-target content. Our method achieves superior editability and visual fidelity, particularly in tasks requiring large-scale shape replacement.

🔥 News

• [2026.1.25] 🎉 Our paper is accepted to ICLR 2026!
• [2025.8.11] Code released!
• [2025.8.11] Paper released!

📖 Pipeline

🛠️ Code Setup

The environment of our code is the same as FLUX, you can refer to the official repo of FLUX, or running the following command to construct the environment.

conda create --n FollowYourShape python=3.10
conda activate FollowYourShape
pip install -e ".[all]"

We recommend you to run the experiment on a single A100 GPU.

🪄 Edit Your Own Image

Toy tests

Gradio demo for image editing will be released soon.

For now, we provide several toy test examples in src/toy_test. You can either run the provided bash script directly or create your own custom bash scripts.

Command Line

You can also run the following scripts in the terminal to edit your own image.

cd src
python edit.py  --source_prompt [your source image prompt] \
                --target_prompt [your editing prompt] \
                --guidance 2 \
                --source_img_dir [the path of your source image] \
                --num_steps 15 --offload  \
                --front [typically set to 1 or 2] \
                --inject [typically set to 3 or 4] \
                --name 'flux-dev' --offload \
                --output_dir [output path] \
                --controlnet_type [specify your controlnet type] \

Please refer to the paper for the rationale and recommended values of the hyperparameters.

🖋️ Citation

If you find our work helpful, please star 🌟 this repo and cite 📑 our paper. Thanks for your support!

@article{long2025follow,
  title={Follow-your-shape: Shape-aware image editing via trajectory-guided region control},
  author={Long, Zeqian and Zheng, Mingzhe and Feng, Kunyu and Zhang, Xinhua and Liu, Hongyu and Yang, Harry and Zhang, Linfeng and Chen, Qifeng and Ma, Yue},
  journal={arXiv preprint arXiv:2508.08134},
  year={2025}
}