Skip to content

The Pytorch Implementation of ''Target-aware Image Editing via Cycle-consistent Constraints''

License

Notifications You must be signed in to change notification settings

HKUST-LongGroup/FlowCycle

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

12 Commits
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

FlowCycle

FlowCycle is a target-aware image editing method with high source consistency.

Target-aware Image Editing via Cycle-consistent Constraints

Yanghao Wang, Zhen Wang, Long Chen

Device Requirements

It requires a GPU with at least 16GB of memory.

Environment installation

conda env create -f environment.yaml

Activate the Environment

conda activate flowcycle

Access Permission of SD-3-medium on Huggingface

pip install --upgrade huggingface_hub
hf auth login

Enter your access token.

Edit the Example Image with SD-3-medium

python demo.py

Source Prompt

A blue-gray Audi car parked in a grassy area. A white dog sitting on the grass, next to the car. A cat laying on the hood of the car.

Target Prompt

A blue-gray Audi car parked in a grassy area. A Husky dog sitting on the grass, next to the car. A tiger cab laying on the hood of the car.

Citing FlowCycle

If you use FlowCycle in your research or wish to refer to the baseline results published here, please use the following BibTeX entry.

@article{wang2025target,
  title={Target-aware Image Editing via Cycle-consistent Constraints},
  author={Wang, Yanghao and Wang, Zhen and Chen, Long},
  journal={arXiv preprint arXiv:2510.20212},
  year={2025}
}

About

The Pytorch Implementation of ''Target-aware Image Editing via Cycle-consistent Constraints''

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages