Inspiration
The inspiration comes from our group member Damanpreet, who came up with the idea! With a major shift to online learning in all levels of education, I am sure teachers/tutors have issues keeping students' attention or would want to know if the pre-recorded lectures they have made are engaging and effective. Our other group member Joseph mentioned how he does tutoring sometimes and it would be interesting to have something to detect if the student is engaged or not, so this could definitely be applicable in the real world.
What it does
The idea of our web application is that an educator can either have students be live streaming themselves through our application or an educator can upload a video of a student watching a lecture and be returned data on how engaged the student might be in the material being presented. The former so far has been partially implemented. The idea is that an educator can upload their lecture and then the student would be able to view this lecture through our site, which would be assessing the student's actions as they watch the lecture. This data would be based on any yawns caught, blinking, and how long eyes are closed. This info would be amazing for educators so they can be more aware of changes they might need to make or how they can improve upon how material is presented for optimal learning.
How we built it
We used the Flask framework to facilitate the frontend creation and Python on the backend. We utilized libraries such as OpenCV and dlib for face recognition and video processing. We used pre-existing models to distinguish emotions and detect faces since we had a short amount of time and I do not think we could have made and trained a model in that time while also making the web application.
Challenges we ran into
One of the biggest challenges, which I am sure a lot of people can relate to, is coordinating a group in different time zones and with new people! It was interesting trying to schedule and break up tasks. In addition, we are all not as proficient in frontend. We struggled with that a little as we went along in the project, and ended up spending a lot of time on that stuff.
Accomplishments that we're proud of
For everyone in our group, this was our first Hackathon, so I think we would say that we are proud that we put ourselves outside our comfort zones and try to create something new! We all have never tried to do a project in this short of time. Most of us have not dealt with video processing either in a project so that was also fun to tinker with and learn about new libraries we could use in the future.
What we learned
That it is really hard to organize a project in a short amount of time! But we also learned a little about OpenCV and video capturing in Python at the end of the day, even if we did not end up with a fully working project at the end. I think it was interesting to learn about a lot of face detection and emotion predicting through machine learning as we looked up resources for this project.
What's next for Tiredness Detection In Online Learning
If we want to continue working on this project, we would definitely clean up the frontend to make it prettier, perhaps a home page and more navigation to make it user friendly. Eventually we could add the ability for students/educator accounts. It would also be interesting to combine and analyze data from all the videos and be able to show it on the site.
Log in or sign up for Devpost to join the conversation.