[AAAI 2026] FIA-Edit: Frequency-Interactive Attention for Efficient and High-Fidelity Inversion-Free Text-Guided Image Editing
conda create -n FSI-Edit python=3.12 -y
conda activate FSI-Edit
conda install conda install pytorch==2.2.0 torchvision==0.17.0 torchaudio==2.2.0 pytorch-cuda=11.8 -c pytorch -c nvidia
pip install -r requirements.txtPlease refer to PIE-Bench.
Run PIE-Bench
source Infer.shRun Single Image
python run_FIA-Edit.py
source Evaluate.shFlowEdit and Our results are here, the rest of our comparision results can be found FSI-Edit and PnP Inversion.
@article{yang2025fia,
title={FIA-Edit: Frequency-Interactive Attention for Efficient and High-Fidelity Inversion-Free Text-Guided Image Editing},
author={Yang, Kaixiang and Shen, Boyang and Li, Xin and Dai, Yuchen and Luo, Yuxuan and Ma, Yueran and Fang, Wei and Li, Qiang and Wang, Zhiwei},
journal={arXiv preprint arXiv:2511.12151},
year={2025}
}