Saint-lsy/Polyp-Gen-Dataset

Polyp-Gen: Realistic and Diverse Polyp Image Generation for Endoscopic Dataset Expansion

Polyp-Gen is a text-guided full-automatic diffusion-based endoscopic image generation framework for realistic and diverse polyp image generation for endoscopic dataset expansion, as presented in Polyp-Gen: Realistic and Diverse Polyp Image Generation for Endoscopic Dataset Expansion. You can use our model for polyp generation.

Code is available here.

Dataset

This dataset was modified by the original LDPolypVideo dataset.

We filtered out some low-quality images with blurry, reflective, and ghosting effects, and finally select 55,883 samples including 29,640 polyp frames and 26,243 non-polyp frames.

[02/26] We update the download link of the training and test dataset at HuggingFace link

Citation

If you find this work helpful, please consider to star🌟 this repo and cite the following paper:

@article{liu2025polyp,
  title={Polyp-Gen: Realistic and Diverse Polyp Image Generation for Endoscopic Dataset Expansion},
  author={Liu, Shengyuan and Chen, Zhen and Yang, Qiushi and Yu, Weihao and Dong, Di and Hu, Jiancong and Yuan, Yixuan},
  journal={arXiv preprint arXiv:2501.16679},
  year={2025}
}

and the original LDPolypVideo paper:

Yiting. Ma, Xuejin. Chen, Kai. Cheng, Yang. Li and Bin. Sun. "LDPolypVideo Benchmark: A Large-scale Colonoscopy Video Dataset of Diverse Polyps", Medical Image Computing and Computer Assisted Intervention Society, 2021