Attention Detector

• 

  Attention Detector

Inspiration:

Both of us went through a series of months of online learning through the 2020 and 2021 academic school years, and we both agreed that there were some serious deficiencies in the educational online process, chief of which being the inability of teachers to gauge and maintain students' attention.

What it does:

This app uses recent psychology research, focusing on assessing the number of blinks, the direction of gaze, and head movement rate to determine a student's attention span and general paid interest in a lecture.

How we built it:

We utilized HTML, CSS, and Javascript to design the front-end web interface, and we used a combination of React, Mediapipe, and Tensor Flow.js to create our AI software.

Challenges we ran into:

One of the main difficulties with a project like this is maximizing accuracy in a situation with so many variables. We wanted to create software with as much precision as possible to legitimize our idea and help the users in real-life situations.

Accomplishments that we're proud of:

Online learning will have far-reaching impacts in the future, even in the post-pandemic world. We're especially proud of the idea and the possible implications it could have in real-life educational environments. We believe this is a unique idea and that other companies and software developers have yet to consider this serious issue, and that our app could prove a helpful solution.

What we learned:

We've learned to believe in our idea and use that motivation to accomplish more in shorter amounts of time. This is an idea we both thought was important and could actually really help a lot of people, and this belief in ourselves gave us the determination to accomplish a lot more a lot faster.

What's next for Attention Detector:

We hope to implement Attention Detector in popular meeting and online classroom applications, such as Zoom and Microsoft Teams, as well as expand its assessment features to include analysis of auditory stimulus.

Built With

• css
• html
• javascript
• mediapipe
• react
• react-router-dom
• tensorflow.js