A novel machine learning scheme for face mask detection using pretrained convolutional neural network.

Area of Science:

• Computer Science
• Artificial Intelligence
• Public Health

Background:

• The COVID-19 pandemic highlighted the need for effective public health measures.
• Face mask usage is a critical strategy for reducing viral transmission.
• Monitoring mask compliance manually is challenging and impractical.

Purpose of the Study:

• To develop and evaluate a deep learning system for automated face mask detection.
• To assess the performance of a pretrained Vgg16 model for this task.
• To improve the accuracy and efficiency of mask usage monitoring.

Main Methods:

• Utilized a pretrained Vgg16 deep learning model.
• Trained only the fully connected layer of the Vgg16 model to reduce computational load.
• Employed two datasets of face mask images (1484 and 7200 images).
• Applied image augmentation techniques to enhance accuracy, especially on smaller datasets.

Main Results:

• The system achieved 96.50% accuracy on a smaller dataset and 91% on a medium dataset.
• The model correctly predicted mask usage on unseen images.
• The combination of Vgg16 and image augmentation significantly improved performance.

Conclusions:

• The proposed deep learning system offers an effective and efficient solution for automatic face mask detection.
• The approach demonstrates the potential of pretrained models and data augmentation for public health applications.
• This technology can aid in monitoring and enforcing mask-wearing compliance during pandemics.