Skip to content

nauman-akram/COVID19_Images_Classification

Repository files navigation

Images_Classification (PyTorch)

Detecting corona-virus (Covid-19) Infections in Chest X-Ray images through Transfer Learning

Find dataset here: https://drive.google.com/file/d/1-HQQciKYfwAO3oH7ci6zhg45DduvkpnK/view

Using Vgg-16 and ResNet-18 pretrained models from pytorch and finetuning just only fully-connected layer of network at first so that they can learn on the learned filters of ImageNet and detect the difference between infected and normal person's X-ray images.

Vgg-16

Vgg-16 fine-tuned only FC layers


Accuracy curves Loss curves

Confusion Matrix

Test Validation

Vgg-16 fine-tuned Entire network


Accuracy curves Loss curves

Confusion Matrix

Test Validation

ResNet-18

Resnet-18 fine-tuned only FC layers


Accuracy curves Loss curves

Confusion Matrix

Test Validation

Resnet-18 fine-tuned Entire network


Accuracy curves Loss curves

Confusion Matrix

Test Validation

“This repository contains code and results for COVID-19 classification assignment by Deep Learning Spring 2020 course offered at Information Technology University, Lahore, Pakistan. This assignment is only for learning purposes and is not intended to be used for clinical purposes.”