---
license: mit
tags:
  - diffusion-model
  - image-segmentation
  - pathology
  - nuclei-segmentation
  - medical-imaging
---

# DiffMix

This repository contains pretrained semantic diffusion model checkpoints from
[DiffMix: Diffusion Model-based Data Synthesis for Nuclei Segmentation and Classification in Imbalanced Pathology Image Datasets](https://arxiv.org/abs/2306.14132), presented at MICCAI 2023.

## Checkpoints

- `consep-ema_0.9999_010000.pt`: checkpoint trained for the CoNSeP dataset.
- `glysac-ema_0.9999_010000.pt`: checkpoint trained for the GLySAC dataset.

Use each checkpoint with the corresponding scripts in the [official implementation](https://github.com/hvcl/DiffMix), specifically `semantic-diffusion-model/scripts_ddim_sample/`.

## Citation

```bibtex
@inproceedings{oh2023diffmix,
  title={Diffmix: Diffusion model-based data synthesis for nuclei segmentation and classification in imbalanced pathology image datasets},
  author={Oh, Hyun-Jic and Jeong, Won-Ki},
  booktitle={International conference on medical image computing and computer-assisted intervention},
  pages={337--345},
  year={2023},
  organization={Springer}
}
```