Facial Recognition and Attendance System Using Dlib and Face_Recognition Libraries

Abstract

An Overview to the computer vision algorithms used for facial recognition. The main article idea is to explore an algorithm which can be used in biometric attendance systems with suitable methods and available inputs. The algorithm mainly uses histogram-oriented gradients for finding the faces, estimation of face landmarks, support vector machine to recognize the face and Deep convolutional networks to compare faces. The basis and scientific procedure for facial recognition is described in the article. A basic application is also developed in order to mark the time of the face appears is .csv format and attendance is marked. The article mainly uses dlib and face_recognition libraries in order to provide the functionality.

Keywords: dlib, face_recognition, DCNN, SVM, HOG, machine learning, face landmark estimation, Attendance system, Computer Vision