AI-Powered Mask Uncovering in Complex Occluded Environments

Abstract

The widespread use of face masks, especially during global health crises such as the COVID-19 pandemic, has posed new challenges to automated facial recognition and surveillance systems. Detecting the presence or absence of a face mask under occluded conditions—where portions of the face are obscured—requires robust and efficient computational techniques. This paper presents a real-time system for uncovering or identifying faces in occluded conditions, particularly focusing on face mask detection and classification. The proposed system leverages a lightweight Convolutional Neural Network (CNN) architecture, MobileNet, combined with OpenCV for real-time video processing, and is implemented using Keras and Python. The core contribution lies in the use of transfer learning, which enhances the system's accuracy and generalization by fine-tuning a pre-trained model on a custom dataset containing both masked and unmasked facial images. The system demonstrates a high degree of reliability in differentiating between covered and uncovered faces even under varying lighting conditions and occlusions. With a focus on computational efficiency and deployment readiness, the model ensures real-time performance without sacrificing detection accuracy. Its lightweight nature makes it suitable for deployment in public surveillance systems at high-traffic areas such as airports, shopping malls, and transportation terminals. This work not only contributes to public health monitoring but also serves as a foundation for future developments in facial analysis under occluded conditions.