Monitoring Student Cheating Activities using YOLOv8

Higher education has had tremendous growth and is starting to transition into the digital era 5.0. It’s critical to identify any unusual behavior by students, especially when it comes to student fraud. This is to guarantee graduates who are of high caliber. Although some study has been done on the recognition of cheating activities, more work needs to be done in this particular scenario because there hasn’t been much done in the way of image quality in particular. Furthermore, it is necessary to enhance the inherent precision of diverse items within the class to enable the system to adjust to a range of environmental conditions. As a result, our research creates an online tool that may identify instances of dishonesty. The YOLOv8 approach is favored for its quick identification speed and excellent accuracy. Results from testing on a database of cheating behavior recognition showed a $97 \%$ accuracy rate. This work advances the development of a web-based tool for promptly identifying instances of student cheating.