Skip to content

Latest commit

 

History

History
34 lines (23 loc) · 2.02 KB

README.md

File metadata and controls

34 lines (23 loc) · 2.02 KB

Real-time Medical Mask Detection during the COVID-19 Pandemic

Note: if you find any problem while running the code, please open a new issue on github. It will be very much appreciated!


This project constitutes the code of our published paper in INISTA 2022, "A Review and Implementation of Object Detection Models and Optimizations for Real-time Medical Mask Detection during the COVID-19 Pandemic", which can be found here and here. Please read it for a more detailed analysis of our results.

  • The models included in this project are written in PyTorch and evaluated on the COCO 2017 dataset and the PWMFD medical mask dataset.

  • The project implements the two following tasks:


Evaluation of different object detection models (Faster R-CNN, Mask R-CNN, RetinaNet, SSD, YOLO) for real-time detection on the COCO 2017 dataset.

Real-time detection model of the correct use of medical mask on human faces based on YOLOv5 using the PWMFD dataset.

Asset files are found at object-detection-assets

Model sources are Torchvision, YOLOv3, YOLOv5, and YOLOv4 repositories.