Inspiration
Schools and Universities across the world have been shut down as a response to the COVID-19 outbreak. As a result, all the procedures have been modified to be conducted remotely. One of the major subsequently arising concerns has been of devising a means of assessing the students on their course material remotely, while also ensuring the system is fair and does not give undue advantage to any of the students. The major obstacle with this regard has been a lack of resources that enabled the supervision of all students to ensure no external aid is used by a student and all students attempt an examination under similar conditions.
What it does
Our team has developed a solution to tackle this issue. We have come up with a design that would enable automated proctoring of a student writing an examination remotely. Our design is in the form of a software that ensures that students remain in front of the computer. This is achieved by tracking the eyes of the students. The software will be designed to accommodate washroom breaks and small snack breaks. The key differentiating factor of our design is that it does not capture videos or photographs, thus ensuring the privacy of students.
How I built it
ExamEye uses the Open Computer Vision library on Python to track an exam-taker's eyes to determine whether they are using unauthorized aids. The program allows for specifically timed-breaks which can be intiated using the web portal at link
The portal works by using the buttons to update a file with the number of break-minutes. This file is downloaded by the python program and the program pauses for the specified number of minutes, allowing the exam-taker to take a break. After the break, the program calls a script located on the ExamEye server which resets the file with a value of 0, thereby resuming the exam.
Challenges I ran into
Python and web integration was something that none of us had experience with. Additionally, our host of choice, InfinityFree was down when we were trying to set it up. Because of this, we had to use an alternate host to place our web portal(000Webhost). Using this was somewhat more complicated than InfinityFree.
Accomplishments that I'm proud of
Our first time integrating a python script running on a local machine with a server on the web was successful! This opens up a world of possibilities for interacting with information on the web, that constantly changes.
What I learned
We learned one way to integrate code that is running on our machines with information available on the web. We also learnt how to scramble and pivot in case of unforeseen circumstances. In our case, one of tea member had to attend to urgent family matter. This member was instrumental in implementing and debugging the main code and was the only one proficient enough to create programs with Open Computer Vision in Python. Another one of our members had to make sure the code worked and integrate it with our web server.
What's next for ExamEye
This project has the potential to be adapted to a multitude of applications. Using its eye-tracking ability, the software can be further developed to collect data about the seating posture of users. This data can subsequently be used to provide medical advice and improve the health of the users.
Another major potential application is by developing the software to analyze and approximate the best locations for online advertisements, on the basis of the movement of the eyes of the user. This can also be utilised to optimize the placement of important content across websites.
Log in or sign up for Devpost to join the conversation.