On 31 December 2019, COVID-19 started in china and then spread all over the world, in which

lockdown to all businesses was obligatory which included educational facilities. That lead to blocking the

teaching/learning process from continuing. To prevent businesses from shutting down, we had to adapt

to the current situation by using modern technology to try and simulate work/learning environments from

home so we used online learning to try to accomplish that. That included online exams which sparked

a new big problem with the amount of students that took advantage from that current situation and the

lack of proctors to cheat their exams and pass their subjects. BirdEye is an application that focuses on

preventing students from cheating online exams with no need of proctors during the exam which will

suit the current lockdown situation. This process was accomplished by using the latest deep learning

technologies.

Our system aims to prevent cheating during exams, this will be achieved by developing a user

friendly desktop application that will have some models that prevent cheating ,the system will

use model that works using deep learning approach with different architectures and preprocessing

which will help the system to have high accuracy to detect cheaters ,our second objective to[4]

have system that work faster than previous systems by solving network problems and user interface

problems, our third objective to have a good server that can handle many student to take the exam.

By the end of the project the system goals:

                  • Environment analysis

                   • Eye tracking model to detect how long the student is looking in different areas away from camera

                   • Students can’t access their courses using a different browser.

                   • The system shall recognize the students faces and voice to confirm their identity.

                   • The system shall block any input or output device that is connected to the student’s PC to prevent them from connecting to                    external monitors or devices that can be used in cheating.

                   • The system shall recognize if there’s any mobile phone nearby that can be used for cheat-ng.[5]

                   • The system shall prevent the student from switching tabs as they can be doing so to cheat.[6]

                   • The system shall report to an admin if there is a student cheating to review the case and finalize if the student is really cheating or                    not.

The system scans the face to identify the identity of the student and whether it is the assigned

student for the exam. It recognizes the students’ voice and detects whether there is any external

and abnormal sounds that might aid the student in cheating.It scans the student’s face and body

gestures and detects any suspicious gestures.It can also detect whether the student is using a mobile

device to cheat or not. Additionally, it can control the student’s computing device to prevent the

students from using basic functions that might help them to cheat (ex: copying, pasting,switching

windows, using an external browser to cheat the exam, etc. . . ).If any cheating activity is detected,

a recording of the cheating incident is stored. A petition option will be available for the cheating

students and if a student petitions the incident, the recording will be sent to an admin to review it

and confirm the officially of the incident