With the rise of COVID-19, wearing masks in public areas has become paramount to preventing the transmission of the coronavirus, Furthermore, many public service providers may require customers to properly wear masks in order to be served. But current enforcement measures of mask mandates are mostly monitored by humans, and thus may be hard to execute effectively in densely populated and fast-moving venues like airports and public transit. An automated mask detection monitoring system could greatly reduce the required human labor. In this study, a facial mask detection software model, usable in existing surveillance applications such as airport monitoring systems, is developed using Keras, a high-level deep learning API, and TensorFlow, an end-to-end open source platform for machine learning. The model uses deep convolutional neural networks (DCNN), which were trained from a database of roughly 12,000 images of faces, with roughly a 50-50 split of masked and unmasked, with the goal of achieving binary classification to detect whether a person is wearing a mask. The DCNN mask detection software achieved a validation accuracy of 98 percent. We also used computer vision techniques from OpenCV including facial detection and contours to detect the location of the people’s faces. The DCNN was then paired with this application to achieve live feed object detection by first detecting a person, and then determine whether they are wearing a mask. Finally, we explore the potential implementation and applications of the facial detection model via the linkage of cameras in monitoring areas to signal and notify officials.As more countries are developing mask wearing regulations, automated masked face detection is a key real-world application [3]. Transportation systems are one area that have been considerably affected by the pandemic, as cities around the world had to enforce massive restrictions on public transport in order to limit transmission of the virus and ensure the safe passage of key workers during the emergency response [4]. Masks have since been required on planes, buses, trains, and other forms of public transportation traveling into, transportation hubs such as airports and stations [5].However, Machine Learning and Deep Learning can help to combat Covid-19 in many ways. Enabling researchers and clinicians to evaluate a vast amount of data to make various predictions. In particular, face mask detection may prove to be useful as a guide for government surveillance on people in public areas susceptible to transmission, to ensure that face mask regulations are upheld. To achieve this, deep learning binary classification facial mask model based on VGG19 and one-shot object detection methods will be trained.We contribute a dual face-and-mask detection system that can provide immediate feedback in live surveillance systems. The method introduced is fast and accurate and can work for live mask-checking in busy public areas, alleviating some of the human effort and error typically subject to the mask mandating task.


    Zugriff

    Zugriff prüfen

    Verfügbarkeit in meiner Bibliothek prüfen

    Bestellung bei Subito €


    Exportieren, teilen und zitieren



    Titel :

    A Real-Time Deep Transfer Learning Model for Facial Mask Detection


    Beteiligte:
    Zhang, Edward (Autor:in)


    Erscheinungsdatum :

    2021-04-20


    Format / Umfang :

    3708604 byte





    Medientyp :

    Aufsatz (Konferenz)


    Format :

    Elektronische Ressource


    Sprache :

    Englisch



    Facial character detection in in real time

    Jha, Preeti / Joshi, Bhavesh | IEEE | 2017


    Real-Time Aerodrome Detection Using Deep Learning Methods

    Koopman, Cynthia / Gauci, Jason / Muscat, Alan et al. | IEEE | 2022


    Real-time traffic incident detection based on a hybrid deep learning model

    Li, Linchao / Lin, Yi / Du, Bowen et al. | Taylor & Francis Verlag | 2022


    Facial Recognition based on Deep Learning

    Aanchal / Nijhawan, Rahul / Goel, Silky | IEEE | 2021


    Discriminative Deep Face Shape Model for Facial Point Detection

    Wu, Y. / Ji, Q. | British Library Online Contents | 2015