# 🧠 DermaVLM Organization

**DermaVLM** is an open-source research organization focused on **resource-efficient medical Vision–Language Models (VLMs) for medicine (our use case is dermatology)**.  
Our work combines **synthetic data generation**, **LLM/VLM training**, **model inference**, and **clinical image annotation tools** into a unified research ecosystem.

This organization is built around the **SCALEMED framework**, enabling scalable and cost-effective dermatology-focused AI research.

You can reach main codes at https://github.com/DermaVLM

## 🧪 Research Foundation

> **📄 Research Paper**  
> This organization is part of the SCALEMED framework.  
> **Read our paper on medRxiv**:  
> **Resource-efficient medical vision language model for dermatology via a synthetic data generation framework**  
> https://www.medrxiv.org/content/10.1101/2025.05.17.25327785v2  
>
> If you use any of our repositories in your research, please cite our paper (see the [Citation](#-citation) section).

## 📖 Citation

If you use this organization or any associated repositories in your research, please cite:

```bibtex
@article{yilmaz2025resource,
  title={Resource-efficient medical vision language model for dermatology via a synthetic data generation framework},
  author={Yilmaz, A and Yuceyalcin, F and Varol, R and Gokyayla, E and Erdem, O and Choi, D and Demircali, AA and Gencoglan, G and Posma, JM and Temelkuran, B},
  year={2025}
}
```

Feel free to explore the repositories, open issues, or contribute to advancing medical-focused small and succesful vision-language models.