Algorithmia Blog - Deploying AI at scale

HackPoly Spotlight: Helping Hand Uses Facial Recognition To Automate Tasks

Students at the HackPoly hackathon

We joined over 500 student hackers at the annual HackPoly hackathon at Cal Poly Pomona last weekend to see what these up-and-coming technologists could develop in just 24 hours. Student developers, designers, and hardware enthusiasts came from all over Southern California to form teams and build innovative products that solve real-world problems using a variety of tools, including Algorithmia.

John Pham, Josh Bither, Kevin Dinh, and Elijah Marchese worked together as team Helping Hand, which focused on creating a platform to give users the ability to remotely automate tasks, such as opening a door based using facial recognition software. Check out the great promo video they made about their project:

We chatted with John Pham to get a closer look at what they built:

How did you build this hack, and what technologies did you use?
This hack uses a Raspberry Pi at its core, responsible for most of the computation done. We then used a master/slave setup involving the Pi and an Arduino Uno. We utilized Algorithmia’s, Microsoft’s, and Clarifai’s API to create a facial recognition technique for a multitude of potential applications. We repurposed a Logitech webcam a live feed to the Pi. We then can programmed the Arduino to activate a servo. We created a SQL database, filled with recognized faces, which the user could personalize. Notifications and logs can be viewed from the Android app, and by extension the Pebble watch.

What’s next for the Helping Hand team?
We all hope to continue refining Helping Hand in propelling the project forward to reach its full potential. As of now, we plan to improve the software aspect of Helping Hand, and then direct our attention on bettering the hardware. We fully intend to release our product to the public. Affordable and user-friendly, Helping Hand will provide the security and peace of mind we all want for our communities and our families.

For John, this was his fourth hackathon working on a completely new platform. Kevin also had prior hackathon experience, but was especially proud of all they were able to accomplish in just 24 hours. According to Kevin, the event was really stressful at the start, because the team hadn’t worked together before, but “it got better as it went along, and we got more acclimated to cooperating.”

HackPoly was Josh’s first hackathon, but like a seasoned hacker, he said “My favorite part of HackPoly were the very early morning hours, from 12 – 3 am, where trying to code coherently becomes nearly impossible. I got about 3-4 hours of sleep for the whole event.”

It was also Elijah’s first hackathon and he described the experience as “a wild rollercoaster ride.” Following the event, Elijah went on to explain that the hackathon was a high pressure, highly competitive endeavor: “There was a lot of stress trying to learn so much in so little time. Despite only having three hours of sleep, Helping Hand turned out great and the satisfaction of completing the project made me feel fulfilled.”

As the winners of the “Best Use of Algorithmia API” prize, we sent the team Cloudbit Starter Kits to help them continue on their path of hardware and Internet of Things hacking. We can’t wait to see what else the team builds and what happens next with Helping Hand!

More About HackPoly and Helping Hand: