Deep Learning Technique for a Identify TE Student System by Face Recognition

Abstract

In this study, we aim to measure the face recognition algorithm robustness of CiRA CORE software and identify a group of TE student with face recognition using the deep learning technique from CiRA CORE software. The one-shot problem and similar-looking face problem are robust performance tests of CiRA CORE. Thus, the number and properties images of a dataset are considered as the robustness of the CiRA CORE software. Identify TE student experiment; three detected faces that are the maximum number of faces in one frame is a condition of the experiment. The results of identify TE student experiment demonstrate that the system gains ninety percent accuracy for the camouflage test by wearing a cap and achieves one hundred percent accuracy for other camouflage problems and identify TE student test. The result of robustness and identify TE student experiment confirms face recognition algorithm CiRA CORE software is effective and robust for identify TE student system.