Remote Exam Integrity through Advanced AI and Camera Vision Techniques

Education stands as the foundation for individual growth and the advancement of society. In this regard, e-assessment

has become an essential and adaptive approach for evaluating students in a remote setting. However, this mode of evaluation

necessitates advanced solutions to uphold the authenticity and fairness of the testing process. Key obstacles include preventing

unauthorized support, maintaining student engagement, and ensuring that assessments accurately measure genuine

understanding and skill.To address these concerns, this project introduces an intelligent, AI-enhanced remote proctoring system

that incorporates computer vision capabilities to establish a reliable virtual examination environment. The backbone of this

system is built on deep learning techniques, especially Temporal Convolutional Networks (TCNs), which analyze sequences of

video frames to interpret user behavior and recognize signs of disengagement or malpractice.The platform integrates various

components such as live webcam monitoring, background audio analysis, and activity tracking of keyboard and mouse inputs.

These interconnected features collectively validate the student's identity and detect any suspicious actions during the test.By

leveraging artificial intelligence for automatic supervision, the system reduces the need for human invigilators while enhancing

the transparency and reliability of digital exams. Its modular, scalable, and efficient design seamlessly aligns with the expanding

ecosystem of online education tools, ensuring continuous learning.