For the ElementX Team, hackathons are one of the company events we look forward to the most.
Over the past weekend, the team gathered together for our third Hackation Retreat, where we have a hackathon within a getaway vacation.
It was the ultimate bonding and learning experience for our team - check out what we got up to!
After our usual afterwork Friday drinks, the team convoyed our way to an out-of-Auckland retreat. We went into this trip with few expectations - we just wanted to see what could be created within 24 hours. All we knew was that we were in for a weekend of great food, amusing board games and potentially the first steps towards disruptive innovation.
We had a family meal together and got in some well-deserved R&R. A few rounds of foosball, pool and board games (Captain Sonar FTW) got our competitive spirit going for the next day’s Hackathon!
After the banter and board games, we gathered the next day for a team brainstorm to exchange any and every idea we had bubbling away in our minds - for work and for play! We sought out concepts that could be used in a variety of applications someday, as well as to build upon projects we already have in the works. With a flurry of Post-Its, we set off as individuals and in pairs to tackle a few different projects.
In under 24 hours, we started the foundations of four applications and pitched them to the rest of the team.
Here's what we came up with:
We created an app that allows you to draw using only your fingers and hand. This was made possible with deep-learning models trained for precise hand tracking, that were efficient enough to run on mobile hardware. The same underlying technology can be used for controlling your devices using only your hand (think while in the kitchen), reading sign language, gaming; the possibilities are endless!
Are you a business trying to gauge your customer’s reactions? Or just a humanoid trying to avoid public faux pas?
We experimented with browser-friendly, on-device vision model to register levels of emotions, such as happiness, sadness, surprise and disgust. This could be utilised to support focus groups for gauging customer reaction to new products.
We’re totally over keeping score at the office Pong Arena! Using Google’s Teachable Machine, we worked on an application to automatically score our lunchtime table tennis matches. By gathering audiovisual data of us playing, we segmented and classified the frames and sounds before feeding them into their separate models for training. The aim of the models was to determine the ball's location via object detection, using the different sound signatures. The output of these two models could then be combined to determine our score and winner!
We were able to create the building blocks of a computer vision model that would allow us to detect and classify grocery items. This could have a great number of uses and applications, including checking you have all your shopping at the supermarket and quality control for food production lines.
Want to be at our next hackathon retreat?
Contact us about our open positions and opportunities!