Performance Analysis of Deep Learning Based Video Face Recognition Algorithm

The identities verification in videos has many applications in area of surveillance, social media and law enforcement. The existing algorithms have obtained higher verification accuracies at equal error rate but it is very difficult to achieve higher accuracy at low false accept rate and this has become major research challenge. An efficient video face recognition system has to develop and the performance is carried out to obtain accurate face recognition from video. We propose a novel algorithm for face verification from video signal and MATLAB is used to implement and simulate proposed algorithm. The performance analysis of proposed algorithm is carried out using databases such as YouTube faces and point and shoots challenge.