Rafli Zaki Rabbani, Siscka Elvyanti, Nurul Fahmi Arief Hakim
Shorter exam formats, such as Multiple Choice Questions (MCQs), which are preferable for online exams has amplified vulnerabilities for exam dishonesty. Online exams easily exploited through search engines or current state of AIdriven platforms such as ChatGPT. Existing proctoring approaches, ranging from authentication, facial monitoring to browser lockdowns, remain limited by requirements and evolving cheating strategies. This research proposes a system designed to prevent remote cheating on online exams, particularly in MCQ exams, by requiring no additional software installation and hardware beyond a built-in camera, enforcing gesture-only answering. A dual-detection algorithm combined hand orientation validation with angle features classification, supported by delay buffers to reduce false inputs, with real-time video streaming and processing was facilitated through WebRTC (Web Real-Time Communication). Experimental evaluation demonstrated that the system achieved robust performance and low latency, which maintained approximately 15 FPS (Frames Per Second). Offering a reliable and practical complement to conventional proctoring methods, enhancing the integrity of online MCQ examinations. © 2025 IEEE.
Electrical Engineering Education, Universitas Pendidikan Indonesia, Bandung, Indonesia
Research at a Glance
Register to unlock
Topics & SDG Alignment
Register to unlock
Collaboration
Register to unlock
Author Profile (Selected)
Register to unlock
References Overview
Register to unlock
Journal & Source
Register to unlock
Metadata & Integrity
Register to unlock