Camera Cheating Behavior Detection System
| Title of the Project |
Camera Cheating Behavior Detection System |
| Students Details |
202111117 Omar Issam Makhlouf |
| Abstract |
This malpractice plague in exams continues to rage on in educational institutions despite a continued high level of integrity challenge. It has been demonstrated that traditional systems of invigilating examinations are incapable, because they are flawed, firstly, by inherent flaws in the system design, second, with inherent biases of the human being, and third, in resources available at a certain time. This current report provides our new solution, the Camera Cheating Behaviour Detection System which is based on AI for identification of behavioural patters that are suspiciously observed and thus related to cheating. Our solution approach focuses on motion patterns and body orientations unlike the use of face recognition techniques, which makes privacy paramount on students while taking their examinations. The machine learning based AI system was developed and trained on behind the complex compiled dataset that represents cheating behaviours emulated by our project members while working on the project. Additionally, a web based exam management system has been developed to provide live color-coded alerts when suspicious activities are detected. Future works include expansion of the dataset to improve detection abilities further, more advanced functionalities for the system itself and compatibility with currently deployed classroom CCTV systems. Our greatest war is on a surveillance system that is efficient, effective and above all impartial to ensure academic integrity within learning boards. |