Zero Exam Fraud and 70% Faster Student Check-Ins: How Face Recognition Is Securing University Exams

To tackle the growing concern of exam impersonation, a university in Southeast Asia sought a fast and reliable way to verify students' identities before exams. Their goal was quite simple: ensure that no one could take the exam on behalf of another.

The university needed a system that would not only keep outsiders out but also speed up the admission process while integrating effortlessly with their existing digital infrastructure.

By the time they reached out to us, they had already developed a concept and created a prototype based on 3DiVi Face SDK.

Flawless Verification Accuracy: The solution had to guarantee rock-solid reliability with minimal false positives, completely eliminating the risk of impersonation.

Integration with university systems: It was crucial that the system fit effortlessly into the university's existing digital infrastructure without disruption.

Data security: Protecting students' biometric data and preventing any potential data breaches were critical.

Scalability: The system had to be flexible enough to adapt to the unique needs of different educational institutions, scaling as required.

To ensure secure identity verification before the exam, we implemented a simple yet reliable solution using 3DiVi Face SDK.

On the server side, biometric templates were created from students' digital photos and then encoded into QR codes before the exam. Since each biometric template is a small binary file (under 300 bytes), up to 9 templates could be stored in a single QR code. The QR code was then printed on the answer sheet, alongside the student's name, and delivered to the exam room.

Upon entering the exam room, students would say their names. The professor or university staff would find the answer sheet, scan the QR code to retrieve the reference biometric template, and then take a photo of the student using a mobile app.

The app would extract the student's biometric template and compare it to the reference. If the templates matched, the student received the answer sheet and was cleared to proceed with the exam.

The solution brought several key advantages:

• No Exam Fraud: Impersonation was completely eliminated, ensuring only the registered student could take the exam.

• Faster Exam Entry: Student registration time dropped by 70%, from 7-8 minutes per student to just 2 minutes.

• Boosted Exam Efficiency: With waiting times to enter the exam room cut by 50%, the university was able to accommodate 30% more students per exam day.

Interested in implementing biometric identification into your project? Book a free consultation and discover how computer vision can solve your business challenges.