In view of the problems of crowded people,foreign bodies may block faces and non-standard wearing masks in public places,YOLOX network is used to identify whether people wear masks,and GIoU loss function is used to improve the detec-tion effect of the model.Albumentations library is used for data enhancement,which can improve the accuracy of capturing faces;GUI design is added to make the user's operation more convenient.The experimental results show that the average accu-racy(mAP)of the model reaches 94.36％,which can accurately identify whether the mask covers the nose and mouth under various environmental factors such as dim light and various obstructions on the face.At the same time,the model has terrific real-time performance and it can be used for testing in classrooms,mall and other public in the future.
关键词
规范佩戴口罩/YOLOX网络/GIoU/GUI设计
Key words
standard wearing mask/YOLOX network/GIoU/GUI design
维普
万方数据