Hand Down, Face Up: Innovative Mobile Attendance System Using Face Recognition Deep Learning

Abstract

Computer Vision is considered as the science and technology of the machines that see. When paired with deep learning, it has limitless applications in various fields. Among various applications, face recognition is one of the most useful real-life problem-solving applications. We propose a technique that uses image enhancement and facial recognition technique to develop an innovative and time-saving class attendance system. The idea is to train a Convolutional Neural Network (CNN) using the enhanced images of the students in a certain course and then using that learned model, to recognize multiple students present in a lecture. We propose the use of deep learning model that is provided by OpenFace to train and recognize the images. This proposed solution can be easily installed in any organization, if the images of all persons to be marked this way are available with the administration. The proposed system marks attendance of students 100% accurately when captured images have faces in right pose and are not occluded.