In higher education, many tasks related to fraud detection have been automated, yet exam monitoring is still conducted by teaching staff and therefore tends to be prone to human error and misestimation. With this project we would like to propose a system based on stereoscopic computer vision that provides a scalable and low-cost approach to automated exam monitoring.
In order to run our code, you will need to install:
Matlab, Matlab Computer Vision System Toolbox, Python 3.6, cv2 and dlib
Link to our web application: http://ids-exam-monitoring.000webhostapp.com/
Play back the output of our incident matching algorithm. We used the left video channel to visualise our results.
Click on entry to review detailed information for each incident including affected frames
Having selected an entry, click again on the top part of the entry in order to jump to the corresponding video position.
In the bottom part, select 'verify' to mark the incident as correct, 'discard' to mark the incident as false and 'cancel' to postpone the decision.