Exam Guard: Enhanced Academic Integrity Surveillance System for Real-Time Behavioral Analysis in Examination Settings

"Exam Guard" is an AI-driven system designed to enhance academic integrity in large in-person exams. Using YOLOv5 for real-time face detection and MediaPipe for gaze tracking, it monitors student behaviors to detect potential cheating, including unauthorized interactions and suspicious head movements. The system alerts proctors in real time, enabling quick interventions, and logs incidents for review and model improvement. "Exam Guard" provides an effective, scalable solution to support proctors and maintain fair examination conditions.