---
license: cc-by-nc-sa-4.0
library_name: diffusers
pipeline_tag: image-to-image
---

This model is a diffusion model trained to generate augmentations for cardiac ultrasound images, as described in [Generative augmentations for improved cardiac ultrasound segmentation using diffusion models](https://hf.co/papers/2502.20100).  These augmentations improve the robustness of segmentation models trained on cardiac ultrasound data.  The model is designed to increase the diversity of training data, leading to better generalization to unseen datasets.

You should cite the following paper when using the code in this repository:

Van De Vyver, Gilles, et al. "Generative augmentations for improved cardiac ultrasound segmentation using diffusion models." arXiv preprint arXiv:2502.20100 (2025).
https://arxiv.org/abs/2502.20100

Blog post: [https://gillesvandevyver.com/#/projects/generative-ai](https://gillesvandevyver.com/#/projects/generative-ai)

Code: https://github.com/GillesVanDeVyver/EchoGAINS

A Jupyter notebook tutorial is available at [tutorial/tutorial.ipynb](tutorial/tutorial.ipynb).