Online Exam Fraud Detection using Different Techniques

This is a real-time face recognition and monitoring system that can be used in applications such as online exam proctoring and security. It uses the face recognition and OpenCV libraries of Python to detect and identify faces through a webcam feed, comparing them to a set of preloaded known face encodings. Recognized faces are labeled generically with ”Student” to maintain anonymity, and any face that the system does not recognize causes an alert after the occurrence of a threshold number of times. An alert is generated in case no face is found for a specific duration, thus continuously monitoring the user's presence. The system works efficiently by resizing frames for faster processing and using the model of HOG, or Histogram of Oriented Gradients, for fast face detection. This approach combines accuracy with speed to enhance fraud detection in online examinations and other similar environments where continuous, real-time face monitoring is critical. Summation forecasts, which are essential for energy planning and management. Index Terms—Face Encoding , frame module, Face detection, Absence detection module.