Please use this identifier to cite or link to this item:
|Title:||E-invigilator using computer vision||Authors:||Tan, Sharlene Shi Yan||Keywords:||Engineering::Electrical and electronic engineering||Issue Date:||2021||Publisher:||Nanyang Technological University||Source:||Tan, S. S. Y. (2021). E-invigilator using computer vision. Final Year Project (FYP), Nanyang Technological University, Singapore. https://hdl.handle.net/10356/149982||Project:||A2094-201||Abstract:||The application of machine learning (ML) techniques in object detection area has been improved drastically in the past decade. The improvements in artificial intelligence (AI) and research in deep learning (DL) and neural networks (NN) has enabled this sector to skyrocket in the late 2010. Object detection merges the task of object classification and localization. Current object detectors are mainly divided into 2 categories: 1. Networks which isolates the location of objects and their classification, such as Faster R-CNN. 2. Networks which predict bounding boxes and class scores simultaneously, such as You Only Look Once (YOLO) and Single Shot Detector (SSD) networks. There are unlimited competitive object detection models which are written annually. The AI that stood out with the greatest number of models designed is computer vision. Ever since mobile applications were introduced, it has garnered an exponential surge in popularity. Currently, almost 80% of people are online through mobile devices and more than 90% of Singapore’s population owns a smartphone.  With the advancement of smartphones and the increasing number of users, it houses so much potential to cater to the needs of different users.The proposed project mainly focuses on the search for the most optimal solution to perform custom object detection to facilitate virtual examination invigilation. A brief study on basic deep learning knowledge, especially the functionalities of various models in this area, are performed. The project conducts an analysis of the different state-of-the-art object detection models, and reasons for settling on using YOLOv4 to design a custom object detection. A mobile application is also developed on Android Studios to create a graphical user interface to perform real-time custom object detection. The application will demonstrate the basic features of E-invigilation such as custom object detection and submission of candidates’ information for authentication. The backend development for authentication is also implemented for examiners' use.||URI:||https://hdl.handle.net/10356/149982||Fulltext Permission:||restricted||Fulltext Availability:||With Fulltext|
|Appears in Collections:||EEE Student Reports (FYP/IA/PA/PI)|
Files in This Item:
|The project is to develop a software application that serves as an invigilator, i.e. an E-Invigilator. The E-Invigilator can monitor students who are taking test remotely. The application will be based on OpenCV (Open Source Computer Vision Library) which is an open source computer vision and machine learning software library. The features of the application will include the detection of all the mandatory objects such as face, hands and monitor screen. The live object detection model can be trained using deep learning to extract the features of the objects to be detected. The remote examiners will be alerted when unusual objects in the monitoring vision are detected.||5.68 MB||Adobe PDF||View/Open|
Updated on Jan 20, 2022
Updated on Jan 20, 2022
Items in DR-NTU are protected by copyright, with all rights reserved, unless otherwise indicated.