Real-Time Biometric Verification and Management System Using Face Embeddings
Ankur Sikarwar, Himanchal Chandra, Indradeo Ram · 2020
With a recent increase in cases of cybercrime, an automated biometric verification system is all the more essential in different organizations for reliable and efficient monitoring. To this end, our paper introduces a scalable and interactive application for biometric authentication using face recognition technology. The application provides four top-level services: Biometric Verification, Show Attendance, User Registration, and User Deletion. We also provide tools for visualization and management of attendance records. The proposed end-to-end pipeline predicts an embedding for the detected face which is then used for face verification based on its Euclidean distance from other face embeddings. On the Face Recognition Data, University of Essex, UK, our proposed method obtains an overall accuracy of 96.76% in the Faces94 and Grimace category.