TRUSTX
Description:
TRUSTX is a smart examination monitoring system that ensures academic integrity in online exams through real-time AI proctoring. It combines OpenCV, Dlib, and YOLOv3-tiny to run six concurrent detection modules — face detection (HOG+SVM), gaze tracking, 3D head pose estimation (PnP), mouth movement analysis (MAR), prohibited-object detection (YOLO), and audio monitoring (RMS) — flagging suspicious behavior in under 500ms per frame on a standard CPU.
Tools:
All the tools that I used in this project:
YOLO Python Dlib OpenCV HOG + SVM EAR MAR NumPy PyAudio imutils Flask MySQL HTML5 · CSS3 · JavaScript PnP RMS NMS + IoU