Adaptive Addax-based deep fuzzy clustering for face mask detection

Respiratory illness Diseases transmitted by droplets highlighted the necessity of face masks. The traditional methods to ensure mask usage faced challenges like the environment, partial masks, low-resolution, real-time processing, and human error, which hinder accuracy and efficiency dynamically. To handle these challenges, efficient automated mask detection systems are required in real-time across various environments. This work develops such an innovative model named Adaptive Addax Optimization Algorithm enabled Deep Fuzzy Clustering (Adaptive AOA_DFC). In this method, preprocessing is done using a Gaussian filter then undergoes Region of Interest (RoI) extraction to isolate the face region. Subsequently, the mask detection is performed using Deep Fuzzy Clustering (DFC) further trained through the proposed Adaptive AOA method. This algorithm is obtained by integrating the Addax Optimization Algorithm (AOA) with an adaptive mechanism to enhance detection accuracy. The performance evaluation metrics are Accuracy, Dice coefficient, and Jaccard coefficient, which yield the values of 92.39%, 0.950, and 0.951, respectively.