Title: Online Proctoring and facial tracking
Team members:
- V Maheysh-19BCE1120
- V.V.Vaishnavi-19BCE1497
- S SuriyaKrishnan- 19BCE1050
Description: Online proctoring and face tracking is useful to authenticate if the candidate is the actual person taking the examination or assessment.
Introduction:
As education is experiencing a tremendous transformation with emerging technologies that allow digital learning, universities and institutions are now looking to conduct entrance & semester-end examinations that allow no compromise in the feasibility, security, convenience and experience of students and educators. An online proctoring system eliminates the difficulties faced for universities during situations like COVID-19 lockdown. Online proctoring occurs when an electronic assessment is digitally supervised and conducted through the internet using the web camera of the test-taker. This can be useful in reducing plagiarism during examinations and human labour.
In our project, we are trying to incorporate a proctoring system, which utilizes a computer’s webcam to detect suspicious behaviours during an online examination and provide feedback to the student. This can be done by detection and recognition of faces, general facial features and tracking and analysis of eye movements. Through face detection, we can find out:
- If there is only one attendee for the exam or if someone is nearby.
- If the person is looking at the screen or elsewhere.
- If the exam taker is absent or too far away from the computer.
- Gaze analysis.
We are applying and implementing both face recognition and eye tracking techniques to complete the functions of proctoring.
The technology to be used will involve Python’s module ‘openCV’ for facial detection, process and analysis. A flask framework (tentative) will be used to integrate the front-end consisting of HTML, CSS and Bootstrap.