-
-
Test Result
-
Homepage
-
Test Page
-
Face Detection Facility
-
Copy Paste Prevention
-
Test Instructions
Inspiration
As a result of the pandemic caused by COVID-19, there is a growing number of institutions choosing to take online examinations while compromising integrity and credibility due to the lack of accessible proctoring facilities. The already available facilities are either lacking in terms of features or are inaccessible to the institutions, which is why we felt the need to develop Shojiki.
What it does
Shōjiki (Japanese: 正直; TL: Honesty) is an application for online examination proctoring with machine vision based monitoring solutions to detect malpractice while requiring minimal human intervention.
Our application also has innovative features such as:
- Our Face Detection algorithm analyses any abnormalities and warns the examinee to abide by the guidelines. If this action is repeated a number of times, the supervisor can be notified of the infringements.
- We also examine the background of the examinee to detect any external assistance of any form.
- It prevents the user from switching windows during the examination.
- It also prevents the users from copying any questions or pasting any answers from the clipboard.
- Our solution also provides a platform for conducting examinations through the application.
How we built it
We built the application using ElectronJS, while the server back-end is built on Flask. For the live camera feed, we are using SocketIO which is then processed real-time with OpenCV. We are using a single shot detection (SSD) model for face detection which is written in Caffe.
Challenges we ran into
Due to high server traffic caused by using SocketIO to send real-time camera feed data, the frame rate had to be reduced to 22 FPS which leads to a delay of roughly 1 second. Due to the time restriction, we weren't able to host the Flask server and currently it needs to be run locally.
Accomplishments that we're proud of
The machine vision algorithm worked as expected, and we were able to get a highly functional and accurate analysis.
What we learned
This was our first time using ElectronJS, as well as SocketIO and we learnt how to build a cross-platform desktop application.
What's next for Shojiki - Online Proctoring Application
We are planning to improve Shojiki into a full-fledged platform for online examination, proctoring and learning management system with features like plagiarism check and in-depth performance analysis for the students as well as the faculty. We are further planning to replace SocketIO with a much more polished WebRTC solution such that the application can be run on a internet connection with lower bandwidth while also reducing the load on the server
Log in or sign up for Devpost to join the conversation.