“Academic Integrity (AI) Guard”
Creators: Elion, Lela, Whaley, Alif, Raymond
High School Students from Midwood High School
Type of Project:
- AI & Machine Learning,
- Data Projects,
- Engineering
Themes of Project:
- Education/Schools
About this Project:
In the United States alone, students on average take around 112 standardized exams throughout their educational career, and so it is clear the significant exposure and impact these tests have on our education. However, a clear and pressing issue in this aspect of education is the excessive instances of cheating that occur during these tests, especially at the University level. We as a group knew that this was an issue we wanted to tackle, as we’re all well accustomed to the standardized test-taking practice, and are well aware of cheating. If built, our project employs the use of both omni-directional engineering and facial recognition Artificial Intelligence programs to help detect classic signs of cheating during exams, and sends alerts to the appropriate proctors.
Project Media
Images
Slides
More About this Project
This project was created as part of an in-person Hackathon in December 2023. Students came up with this idea from scratch in about 2.5 hours!
Share this project:
Twitter
FacebookCommunity Feedback:
Fun
Innovative
Helpful
Great Design
Makes Me Think
Feedback from the Judges:
Your pitch really tells the story of the problem you're trying to solve and how your project will help solve it. I especially love that you included some data in your pitch, that really drives home the impact! I didn't see in the deck what happens if a student is suspected of cheating but I had a thought about someone getting their score thrown out incorrectly. Maybe having a human proctor ultimately making the call on whether or not cheating occurred can help reduce the risk of the AI making a mistake.
Great job in setting up the problem statement and citing data to reinforce it. The 3D renderings of the proposed solution were very cool to see and had some thoughtful details. I think this AI monitored solution is an interesting starting point for the problem of cheating but it mostly raises ethical questions in my mind.
"How might we monitor students without recording personal data?" (e.g. faces, voices)
"How might we reduce bias for neurodivergent students without compromising their personal information?"
"Is consent to be recorded required?"
Your team began to touch on some of these issues but as you push further into this problem and work through them you'll no doubt come away with an in-demand product.
"How might we monitor students without recording personal data?" (e.g. faces, voices)
"How might we reduce bias for neurodivergent students without compromising their personal information?"
"Is consent to be recorded required?"
Your team began to touch on some of these issues but as you push further into this problem and work through them you'll no doubt come away with an in-demand product.
An innovative solution to a very difficult problem to solve. You've clearly defined the problem (50% is very high) and have presented a reasonable solution. I would like to see the ethical aspects of this solution delved into more deeply as even though this could be another tool in the toolkit for detecting cheating it could also be used for many other purposes.