Real Time Face Mask Detection Using Tiny YOLOv3
Adhitya Velip, Amita Umesh Dessai · 2022
To stop the Covid-19 from spreading different techniques were developed and implemented. To make the use of face mask was effective to control the spread of it, as it transmits from one person to other. The use of face mask correctly is also important. The main objective is to develop a technique which can be used to monitor the people wearing face mask, not wearing the face mask or incorrect face mask and also to make use of the model which can be used on portable devices without the use of GPU for real time detection. The Yolov3-tiny model was used for detection. It achieved the mean average precision of 68% and it runs at 13 FPS without the use of GPU. The model was tested for image detection and also for real time detection.