Facilitating Swift Reunions: A Comprehensive Web Application for Missing Children Tracking Using Face Recognition

Every year, countless numbers of children are reported missing. A significant portion of the youngsters in situations of missing children are still unaccounted for. In this research, a unique application of facial recognition-based deep learning technology is presented for recognizing the reported missing kid from a large collection of accessible photographs. In this work, a facial recognition system is designed to accurately identify and track the missing child from the web application’s database where the images of missing children are provided by the public. The Siamese Neural Network is utilized, which ranks similarity between inputs naturally by using a similarity function and is useful in case of smaller datasets containing limited samples per class. A pretrained ArcFace deep architecture based on MTCNN is used to extract face descriptors from the pictures. An average accuracy of 97.34% is observed for 17 different classes.