Development of an automated online proctoring system

The rapid development of machine learning technologies, the increasing availability devices and widespread access to Internet have significantly contributed growth distance learning. Alongside systems, proctoring systems emerged assess student performance by simulating work a teacher. However, despite image processing modern still limited functionality: some not implemented computer vision methods algorithms satisfactorily enough (false positives when working with students different ancestry, racial background nationalities) classification actions (very strict requirements for behaviour), so that software products even refused use modules elements artificial intelligence. It is also problem current are mainly focused on tracking students' faces gaze do track their postures, actions, andemotional state. it assessment emotional state crucial only process itself, but well-being students, as they spend long periods time at computers or other during learning, which has great impact both physical health stress levels. Currently, control over these indicators lies solely teachers oreven themselves, who through test materials independent own. An additional quality storage personal data, most require be identified using identitydocuments store full, unanonymised video servers. Based analysis all problems impede potentially affectstudents' in run, this article presents functional automated online including need analyse human activity monitor hygiene practices process, well maximum protection data. A prototype main components an system meets proposed been developed