Inspiration

Given the state of the pandemic and it's effect on individual conduct, we wanted to introduce a concept that may provide some quality of life improvements for the student population. With the mass transition to online courses, the precedence of mass open online courses(MOOC) serves to highlight the potential drawbacks of online learning, if not apparent already. Even through conventions of common sense, there exists undeniable correlation between attentiveness and learning efficiency. In particular, the downfall of MOOC is the lack of accountability, resulting in lower attention retention, a contrasting reality when compared to that of a traditional classroom setting. Our product proposes the use of eye-tracking and face detection to determine the presence or absence of user attention during online learning. The demo framework was built with Reactjs and the face detection was achieved through the WebGazer.js library.

What it does

Our product uses face detection and eye-tracking software to determine whether or not student engagement is present during a model lecture. The recorded lecture will pause and continue based on the presence or absence of user attention. This is the baseline demonstration of how accountability can be implemented in online learning.

How I built it

We built the model framework using react.js and the software was modified from WebGazer.js library.

Challenges I ran into

Interpreting the algorithms and regressions used in WebGazer.js proved to be the hardest challenge.

Accomplishments that I'm proud of

We are proud that our modification of WebGazer.js works to our baseline requirements. The video player reacts according to our hard-coded indicators of loss of attention.

What I learned

We previously did not have any experience working with WebGazer.js or any form of web development. This assignment proved to be extremely fruitful as we learned a lot in terms of web developement, html, and css.

What's next for Lecture Proctor

Hopefully, the concepts introduced in Lecture Proctor can be extended to other forms of online learning. For one, our baseline model can definitely be improved upon. The existing implementation of a lecture proctor is far from done. Looking even further, this technology can be used in exam proctoring and even live lecture monitoring.

Built With

Share this project:

Updates