Missing Child Identification System using Convolutional Neural Network

Abstract

This paper proposes a new way of using deep learning to use face recognition to find a missing child from photos of other children. People can take pictures of children they think might be up to no good and upload them to a website with landmarks and comments. This picture will match the pictures of children who have been reported missing. The image of the child entered is matched with a photo that is the best match, and that photo is chosen from a database of missing children. With the help of pictures uploaded by the public, a deep learning model is trained to find the right missing child from a database of missing child cases. The Convolutional Neural Network (CNN) is a deep learning method that is very good for image-based applications. This project uses it to recognize faces. With the help of a CNN model with VGG-Face deep architecture that has already been trained, face descriptors are taken from these images. Our algorithm uses a convolution network, which, compared to normal feature extractors, is a high-level feature extractor. Deep learning applications. Child recognition is done based on the trained KNN classifier. Choosing the best performing CNN model for face recognition, VGG-Face, and proper training results in a deep learning model invariant to noise, illumination, contrast, occlusion, image pose, and the age of the child that outperforms earlier methods in face recognition based on missing child identification.