Diabetic_retinopathy_gradcam

A Short Description Of Project

This repo contains Gradcam visualization of Retinal fundus Images from a trained efficientnet_b5 model. Gradcam has been calculated from an intermediate layer which can highlight Fat deposits, Isolated medium sized haemorrhages effectively. For more better visualization Accumulated gradcams calculated from several layers can be very useful

Environment Setup:

Installation instructions

_Run the commands in a terminal or command-prompt.

• Install Python 3.6 or >3.6 for your operating system, if it does not already exist.

• For Mac

• For Windows

• For Ubuntu/Debian

sudo apt-get install python3.6

Check if the correct version of Python (3.6) is installed.

python --version

Make sure your terminal is at the root of the project i.e. where 'README.md' is located.

• Get virtualenv.

pip install virtualenv

• Create a virtual environment named .env using python 3.6 and activate the environment.

# command for gnu/linux systems
virtualenv -p $(which python3.6) .env

source .env/bin/activate

• Install python dependencies from requirements.txt.

 pip install -r requirements.txt

How to Run

Run the script from terminal using following command

 python3 Test_gradcam.py 

Now in the project root directory you will find all the Gauss, Gradcam and Overlapped (Gauss+Gradcam) images.

Contacts

1. Md. Faisal Ahmed Siddiqi (ahmedfaisal.fa21@gmail.com)