aptos-blindness-pytorch

Kaggle's Aptos Blindness Detection challenge solution using pytorch. This repo is a work in progress.

Overview

There is a large set of retina images captured using fundus photography. Training dataset is labelled according to the severity of diabetic retinopathy in following categories.

0 - No DR

1 - Mild

2 - Moderate

3 - Severe

4 - Proliferative DR

Goal is to predict this severity for a new image.

Requirements

following needs to be installed befor running anything :

• torch
• torchvision
• catalyst
• timm
• albumentations

Run

Modify config file in config/config to update parameters according to your needs

For training:

python train.py

Logs are stored under ./logs folder with unique folder name. Tensorboard logs can be plotted as :

tensorboard --logdir ${LOGDIR} --host 0.0.0.0

Results

Training a resnet50 backbone for classification results in :

Accuracy

Loss