Published Paper Details:

Conventional proctoring systems will typically use secure browsers, or plug-ins, or third-party applications, adding technical complexity, restricting user access, and bringing with them many problems. With the growing popularity of online exams, it has become more difficult to keep academic integrity in remote testing. Physical supervision is absent, and this offers a chance to engage in dishonest behavior, which can be in the form of using secondary devices, communicating with other people, or impersonating others. To overcome these obstacles, a multi-modal online exam proctoring system using AI is proposed in this paper. The system employs a collection of AI methods, such as face recognition, eye-tracking, object recognition, sound analysis, and activity monitoring on social media to deliver real-time monitoring and alert suspected behavior. The proctoring system developed based on the Django framework and MySQL to store data and provide constant authentication during the exam with the minimum disturbance to the user. Any violations detected are stored and displayed on a single main dashboard to the examiners, which offers real-time information and reports. This new type of approach makes online testing more secure, provides online learning with a better experience and provides new prospects regarding the use of AI in the education industry.