CXR_C19_DS

Chest X-Ray Covid-19 Detection System

Pre-trained vgg-16 model architecture:

Description

Covid-19 detection system using chest x-ray images based on transfer learning. We used VGG-16 pre-trained model and fine-tuning to adapt model on our issue.

Authors

• Milad Sadeghi DM @EverLookNeverSee

Demo

Run command below to display arguments:

$ python3 demo.py --help

-m    --model          Path to saved model directory
-i    --image          Path to image file
-v    --verbose        Level of verbosity

Example:

$ python3 demo.py --verbose --model <path_to_saved_model_directory> --image <path_to_image_file>

Training on your own dataset

Run command below to display arguments

$ python3 train.py --help

-d    --dataset             Path to dataset directory
-b    --batch-size          Batch size
-l    --learning-rate       Learning rate
-v    --verbose             Level of verbosity
-w    --workers             Number of workers(Hardware)
-e    --epochs              Number of epochs
-s    --save                Specify a directory path to save trained model

Example:

$ python3 train.py -d ~/dataset -b 32 -l 0.00001 -v -w 32 -e 100 -s ~/models/

Useful links

• Dataset: COVID-19 Radiography Database
• Trained model: Download from google drive

License

This project licensed under the MIT License - see the LICENSE file for more details.