Deep Learning-Based Face Mask Recognition System with YOLOv8

A system for recognizing face masks is a technological solution that employs computer vision and machine learning methodologies to detect and ascertain the presence or absence of a face mask on an individual. The utilization of face masks has garnered substantial significance and prevalence in contemporary times, primarily attributable to the COVID-19 pandemic, wherein it has emerged as a pivotal measure to impede the transmission of the virus. The facial mask recognition system is commonly comprised of two primary stages, namely face detection and mask classification. During the face detection stage, the system can detect and localize the existence of a human face within an image or video frame. The categorization of masks can be executed through a range of machine learning methodologies, including deep learning algorithms. The algorithms are trained using a substantial dataset of labeled facial images, which includes both masked and unmasked faces. This training enables the algorithms to discern the distinctive characteristics between the two categories. This work utilizes the You Only Look Once (YOLO) v8 algorithm to discern the presence or absence of a mask on a given subject, subsequently classifying them into two distinct categories based on the extent to which they are wearing a mask, namely “no_mask” and “mask”. The present investigation utilized the Face Mask Dataset as the primary data source for our empirical analysis. The assessments and appraisals of these models incorporate essential criteria. Based on the available data, it can be inferred that YOLOv8s has attained the maximum mean average precision (96.1%) in the “mask” category.