Cheating Detection In Online Exams Based On Captured Video Using Deep Learning
Aysha Sultan Alkalban, Mohammad Mehedy Masud
كلية تكنولوجيا المعلومات-جامعة الإمارات العربية المتحدة · الامارات
الموضوعات
علوم تطبيقية وتكنولوجية
الملخص
Today, e-learning has become a reality and a global trend imposed and accelerated by the COVID-19 pandemic. However, there are many risks and challenges related to the credibility of online exams which are of widespread concern to educational institutions around the world. Online exam system continues to gain popularity, particularly during the pandemic, due to the rapid expansion of digitalization and globalization. To protect the integrity of the examination and provide objective and fair results, cheating detection and prevention in examination systems is a must. Therefore, the main objective of this thesis is to develop an effective way of detection of cheating in online exams. In this work, a system to track and prevent attempts to cheat on online exams is developed using artificial intelligence techniques. The suggested solution uses the webcam that is already connected to the computer to record videos of the examinee in real time and afterwards analyze them using different deep learning methods to find best combinations of models for face detection and classification if cheating/not cheating occurred. To evaluate the system, we use a benchmark dataset of exam videos from 24 participants who represented examinees in online exam. An object detection technique is used to detect face appeared in the image and crop the face portion, and then a deep learning based classification model is trained from the images to classify a face as cheating or not cheating. We have proposed an effective combination of data preprocessing, object detection, and classification models to obtain high detection accuracy. We believe that the suggested invigilation methodology can be used in colleges, institutions, and schools to look for and keep an eye on suspicious student behavior. Hopefully, by putting the proposed invigilation method into place, we can aid in eliminating and reducing cheating incidences as it undermines the integrity and fairness of the educational system.
روابط وملفات
التعريف والنوع
- رقم الوثيقة
- edb1b619-98b2-458e-854a-52bc6789989c
- رقم العقد
- 0
- نوع الوسائط
- Crawler
- نوع المحتوى
- الرسائل العلمية
- صيغة المصدر
- رسائل ماجيستير
- نوع الملف
- pdf text
- أسماء الملفات
- 2325482_2.pdf
بيانات النشر
- ألقاب المؤلفين
- [{"name_ar":"Aysha Sultan Alkalban","title_ar":"اعداد","title_en":"Preparation"},{"name_ar":"Mohammad Mehedy Masud","title_ar":"اشراف","title_en":"Supervision"}]
- اللغة
- English
المصدر والدورية
- اسم المصدر
- Cheating Detection In Online Exams Based On Captured Video Using Deep Learning
المحتوى والصفحات
- عدد الصفحات
- 0
إشراف وإعداد
- الإشراف
- Mohammad Mehedy Masud
- الإعداد
- Aysha Sultan Alkalban
الاقتباسات الببليوغرافية
APA
MLA