NotesFAQContact Us
Collection
Advanced
Search Tips
Back to results
Peer reviewed Peer reviewed
Direct linkDirect link
ERIC Number: EJ1433271
Record Type: Journal
Publication Date: 2024-Jul
Pages: 15
Abstractor: As Provided
ISBN: N/A
ISSN: ISSN-1360-2357
EISSN: EISSN-1573-7608
Video Based Action Detection for Online Exam Proctoring in Resource-Constrained Settings
Dilky Felsinger; Thilina Halloluwa; Ishani Fonseka
Education and Information Technologies, v29 n10 p12077-12091 2024
Academic misconduct is a growing problem in online education. While there are ways to curb academic misconduct in online exams, utilization of technology to proctor online exams in a simple manner in limited-resource settings remain unclear. This study set out to identify a reliable technique for utilizing webcam footage to identify instances of academic dishonesty. Instead of applying component-based feature extraction, this study provides a deep learning-based approach to online exam proctoring that recognizes exam candidates' behaviors directly from video input. By using an online exam dataset consisting of twenty-four test-takers who replicate real-world actions, experimental results show the efficiency of the approach. The article provides a collection of webcam recordings with annotated actions to evaluate the proposed approach. The findings show that a deep learning model using the Slow Only variation has a true detection rate of 78.9%. The visualization module reduces the amount of time invigilators must spend watching videos to capture academic misbehaviors by offering a comprehensive graphical view with details of a candidate's actions during an exam. The study's conclusions will help design an effective and efficient system for online exam proctoring in resource-constrained environments.
Springer. Available from: Springer Nature. One New York Plaza, Suite 4600, New York, NY 10004. Tel: 800-777-4643; Tel: 212-460-1500; Fax: 212-460-1700; e-mail: customerservice@springernature.com; Web site: https://bibliotheek.ehb.be:2123/
Publication Type: Journal Articles; Reports - Research
Education Level: N/A
Audience: N/A
Language: English
Sponsor: N/A
Authoring Institution: N/A
Grant or Contract Numbers: N/A