Model Card for RETFound_MAE_MEH

This modelcard aims to provide a pre-trained vision foundation model RETFound, pre-trained with Masked Autoencoder.

This is the official weight for RETFound Nature paper

Model Details

Model Description

• Developed by: Yukun Zhou
• Model type: Pre-trained model
• License: Creative Commons Attribution-NonCommercial 4.0 International Public License (CC BY-NC 4.0)

Model Sources

• Repository: RETFound
• Paper: Nature paper

Uses

This repo contains the model weight. After granted the access, please fill the token in the code.

The code will automatically download the model and run the training.

Environmental Impact

• Hardware Type: 4 * NVIDIA A100 80GB
• Hours used: 14 days
• Cloud Provider: UCL CS Cluster & Shanghai Jiaotong University Cluster

Citation

@article{zhou2023foundation,
  title={A foundation model for generalizable disease detection from retinal images},
  author={Zhou, Yukun and Chia, Mark A and Wagner, Siegfried K and Ayhan, Murat S and Williamson, Dominic J and Struyven, Robbert R and Liu, Timing and Xu, Moucheng and Lozano, Mateo G and Woodward-Court, Peter and others},
  journal={Nature},
  volume={622},
  number={7981},
  pages={156--163},
  year={2023},
  publisher={Nature Publishing Group UK London}
}

Model Card Contact

ykzhoua@gmail.com or yukun.zhou.19@ucl.ac.uk