Identifying People Wearing Masks in the Wild by Yolov7 Algorithm

Abstract

Rapid face detection is an important matter at the present time, as face detection while wearing a mask has today become an important security matter in institutions, as well as protecting society from the spread of diseases, especially after the outbreak of infection with the Corona virus. Covid-19 virus. Due to the rapid spread of the Covid-19 pandemic, there is a need for society to adhere to wearing a mask in all public institutions, to prevent the spread of this disease. Researchers worked to find solutions to recognize faces and distinguish a person's identity. There was a problem in detecting faces and recognizing them easily, as researchers found many solutions to detect faces. So far, detecting faces while wearing a mask has problems with accuracy. In this research, a model of the deep learning algorithm, YOLOv7, will be used. It is a YOLO model that is characterized by accuracy and speed compared to previous YOLO models, and compared to a deep learning algorithm that performs two-stage detection, such as the CNN algorithm. Here the YOLOv7 model of the YOLO algorithm is proposed for face detection and recognition with and without mask. It is a model in which the structure of the algorithm has been modified, as it is distinguished by its speed in detecting faces compared to its predecessors. also reviewed most of the experiments with algorithm YOLO (You Only Look Once) detection algorithms and CNN (Convolutional Neural Network), and noticed that YOLOv7 is a better model than previous YOLO models in detecting faces while wearing a mask in terms of speed and accuracy. Face detection and discrimination has become very important at the present time from a security standpoint in all public places and requires accuracy and speed in detection.