Skip to content

reshalfahsi/myqlanet

Repository files navigation

Build Status

MyQLaNet

A Deep Learning Platform for Macula Detection.

It provides end to end system for macula detection with graphical user interface.

alt text

Dependencies

  • Ubuntu Linux (OS)
  • PyQt5 (GUI)
  • Python > 3.x (Programming Language)
  • PyTorch (Machine Learning Framework)
  • OpenCV and scikit-image (Computer Vision Framework)

Installation

sudo apt install pyqt5-dev-tools
sudo pip3 install -r requirements.txt --no-cache-dir
python3 app.py

Working with the Library

Instead of using GUI, you can code from the scratch:

from myqlanet import *

# path to the important config file
dataset_path = '/path/to/dataset'
annotation_path = '/path/to/annotation'
weight_path = '/path/to/weight'

# create MyQLaNet model
model = MyQLaNet()

# create dataset
dataset = MaculaDataset(annotation_path, dataset_path)

# training time!
model.compile(dataset)
model.fit(weight_path)