hcarrion/atopic_dermatitis

cgDDI: Textual Inversion for Atopic Dermatitis

This repository contains the textual inversion adaptation weights for atopic_dermatitis (eczema), built on top of stabilityai/stable-diffusion-2-1-base.

These weights were developed as part of the paper: Controllable Generation of Diverse Dermatological Imagery for Fair and Efficient Malignancy Classification (MICCAI 2026).

• Paper: Hugging Face Paper Page
• Repository: GitHub - ControllableGenDDI
• Dataset: hcarrion/ControllableGenDDI Dataset

About cgDDI

cgDDI (Controllable Generation of Diverse Dermatological Imagery) is a hybrid framework designed to address the equitable performance bottleneck in dermatological AI. It enables:

1. Synthesis of realistic healthy skin samples.
2. Non-parametric mapping of rare lesions onto novel skin-tones and locations.
3. Efficient parametric generation of diverse dermatological imagery using textual inversion.

This specific checkpoint can be used to generate synthetic images of atopic dermatitis for fair and robust machine learning model training.

Citation

If you find cgDDI helpful in your research, please cite:

@inproceedings{carrion2026cgddi,
  title     = {Controllable Generation of Diverse Dermatological Imagery for Fair and Efficient Malignancy Classification},
  author    = {Carri{\'o}n, H{\'e}ctor and Norouzi, Narges},
  booktitle = {Medical Image Computing and Computer-Assisted Intervention (MICCAI)},
  year      = {2026},
  publisher = {Springer},
  series    = {Lecture Notes in Computer Science}
}