Recently, with the COVID-19 pandemic shutdown measures introduced by many countries all over the world , universities were forced to move from physical classroom settings to remote classes to help flatten the contamination curve. With this shift, for a selected set of remote classes and especially examinations, it is necessary to check, if the student participating in the class or in the exam is enrolled in the course or not and if he/she is who he/she claims to be (individuals authentication) to avoid frauds. In this sense, to verify such claims, facial recognition using the student’s webcam and a valid identification (ID) card (e.g., the University of Zurich - UZH student card) could be employed.
The idea is to use such a system internally at UZH, such that a necessary database with relevant UZH data can be utilized for certain checks. The thesis plans to run an image/facial recognition test to check if the student is the same as on the photo of the UZH card. With a text extraction, the system checks if the student is enrolled in the current semester. And also with text extraction, the system compares Student-IDs with the student IDs enrolled in the particular course.
Supervisors: Bruno Rodriguesback to the main page