[O] Project OnSite

• 

  Attendance in an instant

Inspiration

One of our primary inspirations behind this project was making the classroom more inclusive for people of all backgrounds. Most will agree that many teachers are not great at pronouncing the names of people from more diverse backgrounds, often leading to awkward situations and embarrassment. In addition to this, everyone will agree that being falsely marked absent is not a great feeling. Therefore, we aimed to fix this problem by creating a method of attendance that is not only more inclusive but increases productivity in the classroom.

What it does

Our project, OnSite, allows teachers to take attendance using a simple camera or webcam by the doorway. As students pass by, using a machine learning facial recognition software, OnSite instantly checks in students to the class, notifying students on their cellphones when they've been checked off. After the class starts, late students are notified of their absence through Twilio. Assuming teachers take around 5 minutes to take attendance, with a 180 day school year, OnSite would save teachers around 14 hours of time on attendance. This equates to about 3 whole school weeks of education (assuming classes are around 1 hour long.) This would all happen with less than an hour of setup on the first day. In addition to this, students can always prove their presence in class by simply showing their text message verification. No more false absences or erroneous tardy marks. Once attendance is collected, a .csv and png file is created in the directory.

How we built it

We built OnSite on Python, using OpenCV for facial recognition, tkinter for GUI, and Twilio for messaging.

Challenges we ran into

One major challenge we obviously ran into was getting the facial recognition running. There were many bumps and logic errors we ran into as we were figuring all this out. Our original plan was to have OnSite be a web application, using Auth0 as our login system and using a .tech domain with Google Cloud. However, the further we got into the project, the more we realized that if we wanted to finish in time, we would have to make some cuts. This was when we pivoted our project towards a less-ambitious non-web based application.

Accomplishments that we're proud of

The thing we're the most proud of over this whole experience is all the hurdles and dead-ends we managed to overcome to reach our end product. While we did have to cut some corners and leave out some features, we all came to something that we're super proud of creating.

What we learned

Even though we are very proud of our project, perhaps the thing we're most proud of is the sheer amount that all of us learned during this 24 hour journey. Every single one of us learned something new during this experience. Whether it be Python, how to make a GUI or Facial Recognition, we all took something from this experience that can be used for years to come.

What's next for Project OnSite

As previously mentioned, we did have to cut some corners on the project in order to meet the deadline. However, with more time, we plan to add a login system for teachers and students, a website in which students would register themselves, thereby taking less in-school time and adding unique IDs per person to avoid weird situations with very similar-looking twins. This is all in addition to further optimizing the application and polishing the GUI.

Built With

• opencv
• python
• tkinter
• twilio