Hugging Face's logo

ClementP
/
FundusDRGrading-vit_large_patch14_dinov2

Transformers
PyTorch
Safetensors
English
fundus
diabetic retinopathy
classification
Eval Results
Inference Endpoints
Model card Files Community
FundusDRGrading-vit_large_patch14_dinov2
File size: 1,524 Bytes 
5a36d11
27a7a78
 
5a36d11
27a7a78
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
5a36d11
27a7a78
 
 
 
 
 
 
 
 
 
5a36d11
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
 
---
language: en
license: mit
tags:
- fundus
- diabetic retinopathy
- classification
datasets:
- APTOS
- EYEPACS
- IDRID
- DDR
library: timm
model-index:
- name: vit_large_patch14_dinov2
  results:
  - task:
      type: image-classification
    dataset:
      name: EYEPACS
      type: EYEPACS
    metrics:
    - type: kappa
      value: 0.7660728693008423
      name: Quadratic Kappa
  - task:
      type: image-classification
    dataset:
      name: IDRID
      type: IDRID
    metrics:
    - type: kappa
      value: 0.7999367713928223
      name: Quadratic Kappa
  - task:
      type: image-classification
    dataset:
      name: DDR
      type: DDR
    metrics:
    - type: kappa
      value: 0.7271304130554199
      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.