---
language: en
license: mit
tags:
- fundus
- diabetic retinopathy
- classification
datasets:
- APTOS
- EYEPACS
- IDRID
- DDR
library: timm
model-index:
- name: efficientnet_b2
  results:
  - task:
      type: image-classification
    dataset:
      name: EYEPACS
      type: EYEPACS
    metrics:
    - type: kappa
      value: 0.7577868700027466
      name: Quadratic Kappa
  - task:
      type: image-classification
    dataset:
      name: IDRID
      type: IDRID
    metrics:
    - type: kappa
      value: 0.7042314410209656
      name: Quadratic Kappa
  - task:
      type: image-classification
    dataset:
      name: DDR
      type: DDR
    metrics:
    - type: kappa
      value: 0.7354801893234253
      name: Quadratic Kappa
---
# Fundus DR Grading

[![Rye](https://img.shields.io/endpoint?url=https://raw.githubusercontent.com/astral-sh/rye/main/artwork/badge.json)](https://rye-up.com)
[![PyTorch](https://img.shields.io/badge/PyTorch-ee4c2c?logo=pytorch&logoColor=white)](https://pytorch.org/docs/stable/index.html)
[![Lightning](https://img.shields.io/badge/Lightning-792ee5?logo=lightning&logoColor=white)](https://lightning.ai/docs/pytorch/stable/)

## Description

This project aims to evaluate the performance of different models for the classification of diabetic retinopathy (DR) in fundus images. The reported perfomance metrics are not always consistent in the literature. Our goal is to provide a fair comparison between different models using the same datasets and evaluation protocol.