Our AI detected this image with reasonable certainty
Our AI detected this image with superior certainty
Our AI did not detect this image correctly
This image returned a very precise and accurate classification
We decided to try out a photo of Susan Ibach from Microsoft and it worked spectacularly!
This one was also very accurate, returning with a reasonable certainty of classification
Our AI couldn't differentiate the team logo in this photo.
We were inspired to create this project after seeing the workshop put on by Microsoft about their new technology that we had not seen before. Azure was new technology to us and we wanted to explore what we could create with this technology. Taking inspiration from the challenges given to us and the large amount of hockey games around our university, we came up with IceBreakers.
What it does
IceBreakers has a trained AI that can tell NHL team logos from jerseys and we added in a way to get game data so you can see how your respective team is doing updated as data comes in. We put this all in a streamlined website so it's easy to navigate
How we built it
We spent a lot of time training our AI so it could differentiate the sports team logos by feeding it tons of images and then we worked with Wix code to create a website so the AI could pull from the uploaded images and run the images against it's knowledge base and also give you real-time game data from the most recent game your team has played in.
Challenges we ran into
We ran into a lot of challenges since we decided to tackle AI. It was an interesting but challenging weekend since we had never worked with AI and half our team had never worked with Wix before. It was difficult to get the AI to cooperate since it was like training a toddler and even AI can throw temper tantrums. We also had trouble getting Wix and our AI to get along since there was a bit of a technological screaming match between them so we had to make them cooperate, and came very close to getting them listen.
Accomplishments that we're proud of
We're proud of the fact that we were able to successfully train an AI, make a web app, and all of this was accomplished in a team of two.
What we learned
We learned that there is a lot that goes into AI and it requires a lot of effort to get a decently trained model. There is also a lot of work to be split up in a smaller team so it really tested our time management and our ambition.
What's next for IceBreakers
We hope to improve the functionality of our web application and to streamline usability and user experience. We also want to incorporate a way to add in the ability to find Facebook groups pertaining to your respective hockey team.