Our project for the Flare Hacks Hackathon.


As the schools were reopening, we thought that it would be nice to create something to help them.

How we built it:

We made our program by first training our own models, and then we tried to look for open-source models that would help us. After that, we developed software for the models, then we added the necessary hardware materials.

Challenges Faced

Our custom trained model required a lot of fine-tuning of the dataset. The computer that we ran the software and hardware on, the raspberry pi, was kind of slow to handle all of the computations needed for the code to run. So, we had to use the NCS stick to handle the computations. The motors did not work at the beginning when we tested it, so it required us to think out of the box to use at-home materials to make it work.

What we learned

  1. How to train a custom model.
  2. How to perform recognition of humans in Computer Vision.
  3. How to play sounds in Python.
  4. How motors can run on the raspberry pi.
  5. Math for detecting the distance between two people.

Google Slides Presentation:

Built With

Share this project: