A Review on Applications of Artificial Intelligence in Online Proctoring Systems
A Review on Applications of Artificial Intelligence in Online Proctoring Systems
Authors:
H.Ashish Kumar¹, V. Siri2, Kosuri Sai Teja3
¹Assistant Professor, Department of Computer Science and Engineering, St. Martin’s Engineering College, Hyderabad, India devasudhacse@smec.ac.in
2Assistant Professor, Department of Computer Science and Engineering, St. Martin’s Engineering College, Hyderabad, India devasudhacse@smec.ac.in
3 Student, Department of Computer Science and Engineering, St. Martin’s Engineering College, Hyderabad, India saitejak.004@gmail.com
Abstract
Online proctoring systems have emerged as a critical component in modern digital education, especially with the rapid adoption of remote learning environments. Traditional examination monitoring methods are not scalable in online settings, creating a need for automated and intelligent solutions. Artificial Intelligence (AI) plays a key role in addressing these challenges by enabling real-time monitoring, identity verification, and behavior analysis.
This paper presents a comprehensive review of AI applications in online proctoring systems, focusing on techniques such as computer vision, facial recognition, gaze tracking, object detection, and audio analysis. It examines how machine learning and deep learning models are used to detect suspicious activities and ensure fairness in examinations. Additionally, the paper evaluates system performance using standard metrics and discusses limitations such as privacy concerns, algorithmic bias, and false positives.
The study further explores future research directions, including explainable AI, multimodal learning, and privacy-preserving technologies. The findings indicate that while AI significantly enhances the efficiency and scalability of proctoring systems, careful design and ethical considerations are necessary to ensure trust and reliability.
Keywords: Online Proctoring, Artificial Intelligence, Computer Vision, Face Recognition, Gaze Tracking, Object Detection, Behavioral Analysis, Cheating Detection, Multimodal Learning.