Leap | Devpost

Inspiration

We are a few College students living together at Olin. We always attend class together, but can't stand having to get up off the couch to point the laptop camera towards us. We wanted something that could automatically find our faces and streamline the zoom meeting world. We also know that having the option to stand up, stretch, and even move around your room while still showing that you are attentive in class can help the hard weeks go by a little faster.

What it does

Our project is an automatic camera-aiming phone holder that uses visual tracking to find the faces of people in the room. In our case, if a lot of people are in the camera shot, the Leaf is trained to find a 'speaker object' that signals which person to focus on. The device always tries to put someone's face in the center of the screen.

How we built it

We did all of our coding in python with the help of OpenCV. All of the parts were planned, designed, and 3D printed during this competition. (A HUGE RUSH). We still have some problems from all the work that entailed.

Challenges we ran into

Organization was one large issue we had. There were some times where people were working on the same thing separately. This was being wasteful of our time, and some could have spent more time moving on, or helping with another part of the project.

Accomplishments that we're proud of

We learned all of the image processing from scratch. Not one of us has used OpenCV before, and we are so proud that we got something to work. We also love that we made something that we very well may use in the future. It wasn't just making for the sake of making. (although there is nothing wrong with that.)

What we learned

We learned that, although we all wanted to explore new areas, its ok to seek help from teammates who already know about it. This would have helped us work a little faster. We also learned about our limits as engineers. Just because we were passionate about the project, that wasn't enough to keep the energy up all night.

What's next for Leap

We want to build a version 2! we think that it would be great to try to build it smaller, lighter, and more robust of course. We really like the program that we came up with, and would like a platform to actually use it in the suite!