This is the repository for "Controllable editing via diffusion inversion on ultra-widefield fluorescein angiography for the comprehensive analysis of diabetic retinopathy".
In the training stage, the CLIP is tuned by prompt tuning strategy using infoNCE loss.
sh scripts/train_clip.sh
In the training stage, the SD model is trained with the multimodal embeddings.
sh scripts/multi_lora.sh
During the inferencing phase, the original UWFA image is edited into disease-free domain.
sh scripts/inference_inversion.sh
Here is an example, which includes the original image, the edited image, and the difference between the two.
If this project is help for you, please cite it.
@article{ma2024controllable,
title={Controllable editing via diffusion inversion on ultra-widefield fluorescein angiography for the comprehensive analysis of diabetic retinopathy},
author={Ma, Xiao and Ji, Zexuan and Chen, Qiang and Ge, Lexin and Wang, Xiaoling and Chen, Changzheng and Fan, Wen},
journal={Biomedical Optics Express},
volume={15},
number={3},
pages={1831--1846},
year={2024},
publisher={Optica Publishing Group}
}