A Smart School by Using an Embedded Deep Learning Approach for Preventing Fake Attendance

In the latest years, deep learning branch becomes the most popular sector of the human life to ameliorate the potentiality of artificial intelligent. In this paper, an embedded smart school system is designed based on face recognition with the use of machine learning and deep learning technologies. The prohibition of fake attempts will be the main objective in the exam hall for specified classes, which is based on face recognition. For this purpose, an embedded platform used to implement an efficient system. There is a face database for specified classes that can be used for finding fake attendance during examination. There might be sometimes someone on behalf of his/her attendance will send someone else with that name to do the recommended test. While entering the students to the exam hall for examination, they will be investigated properly before doing their tasks. Meanwhile, when an anonymous student enters the hall of exam, directly this smart system will alert and emit notifications and an anonymous picture will be sent to the teacher's smartphone, which is based on Internet of Things (IoT). The suggested systems is exact fast and has low computational cost.