Smart Exam Surveillance System
Smart Exam Surveillance System is not just a project it?s a solution to a problem in our society. The problem of using unfair means in an examination. This is basically an automation to the work of an invigilator in an examination hall but with better accuracy. Exam surveillance system involves 3 ma
2025-06-28 16:35:20 - Adil Khan
Smart Exam Surveillance System
Project Area of Specialization Artificial IntelligenceProject SummarySmart Exam Surveillance System is not just a project it’s a solution to a problem in our society. The problem of using unfair means in an examination. This is basically an automation to the work of an invigilator in an examination hall but with better accuracy. Exam surveillance system involves 3 major task including Cheating detection in real time once any malicious activity is detected the next job of the system is to identify the individual that has been caught doing that malicious activity in the examination hall. And lastly the system would automatically interpret the entire scene and alert the management regarding the activity with recorded video streams through a web portal as evidence of the act. As much as the technology is growing providing better tools and technology for high accuracy there still always exists a possibility of an error hence in order to tackle this possibility the exam surveillance system includes a review System so that students get the right to defend themself. To address the major challenge of the project of Human Action Recognition at real time YOLOV3 is being used furthermore We have used pre tested FaceNet Algorithm for face recognition, whereas for web portal Flask framework is used.
Cheating in exams is one of the biggest and most common issues in Pakistan. It is clear that cheating in exams destroys the future of the students. Today this malpractice is embedded in the culture of our educational institutions and a necessary part of the adventures of student life.
In recent years, the number of surveillance cameras installed in classrooms to monitor behaviour of students while examination has been increased drastically . The video streams are constantly recorded and are seen randomly from time to time. But this system is less productive and useful as it is impossible to monitor whole length video streams by humans, sitting in front of a screen. In an ideal system, software would automatically interpret the entire scene and alert in case of any suspicious situation. This would bring more transparency to the system as it will become more robust in nature. So the motivation for our project is that we would automate the work of an invigilator by developing a Deep learning based Exam surveillance system.
The essential point of this project is to utilize the video streams from CCTV cameras in classrooms to its maximum limit and make it as useful as possible. Our Project mainly address Human Action Recognition Problem where the main challenge is to detect action with great accuracy in real time.
Academic dishonesty is a widespread practice, across all age groups. Our system “Smart Examination surveillance System” is planned with the mind set to reestablish the education system of Pakistan where the culture of cheating in exams do not exist and in order to achieve that goal we will be developing a system that implements cheating detection by using human action recognition algorithms and not only it will detect cheating but also will identify the student involved in the act and will notify the authorities and the student involved. The system will be doing it in real time. The systems working flow is initiated first by the CCTV video streams that will work as the input for the YOLOV3 algorithm which is used to recognize the action of the students present in the examination hall and as any illegal action will be detected the software will identify it and then the next phase is face recognition for which FACENET is being used the frames with illegal activity detected by YOLOV3 will then be sent to FACENET as input for face recognition of the student. As the face is detected the system will generate an email to the student detected and will also inform the authorities.
There exists a chance that the system might wrongly detect an activity as illegal hence to tackle that scenario there is a student review system designed as well. The student if receives an email from the system then they have to first request a defense and will be submitting a defense document. Now the decision lies on the exam coordinator to decide whether to accept or reject the document. If it is accepted the system will then remove that record from the cheating cases list. If they reject the authorities then they will take action against it.
Following are the benefits of this project:
- The successful implementation of this project could lead to reduction of cheating in exam culture.
- This system can be deployed in schools colleges and universities to get transparent results.
- This system would eventually contribute to country cause as external world donot accepts pakistani degree, one of its main reason is there faith on our examination system.
The final deliverable of our project would be a Project Report Documentation that will contain the system architecture and ready to deploy system.
Final Deliverable of the Project HW/SW integrated systemCore Industry EducationOther IndustriesCore Technology Artificial Intelligence(AI)Other TechnologiesSustainable Development Goals Quality EducationRequired Resources| Item Name | Type | No. of Units | Per Unit Cost (in Rs) | Total (in Rs) |
|---|---|---|---|---|
| Total in (Rs) | 70000 | |||
| Graphic Card | Equipment | 1 | 60000 | 60000 |
| IP Camera | Equipment | 2 | 5000 | 10000 |