As a group we have a passion for machine learning. Naturally, we were drawn to the Clarifai API for it's powerful image recognition capabilities. We are four health minded youth, who want to help others by making nutrition tracking more accessible, for those on the go.
What it does
Given a picture of food. Our web application allows the user to crop an image, and focus on specific parts of the meal. The app then runs the image through the Clarifai API, in order to acquire the proper tags. We then filter out tags unrelated to food, and use the best fit to perform a query on. The query is performed using the FatSecret API, which returns the user a simple summary of the nutrition facts, but also allows for a more detailed view. In the event that the information produced is incorrect, we offer the ability for the user to manually tag the item. This not only helps improve the user experience, but allows Clarifai (and eatr) to improve over time.
How we built it
We initially split our team into 2 pairs, where each pair worked on a separate API (Clarifai and FatSecret). After this, we cooperated to integrate all of it together with a visually appealing and straightforward front end.
Challenges we ran into
Accomplishments that we're proud of
In the end we are extremely proud of how well we were able to work together as a team towards a common goal. :"). We put in a lot of time for this competition, with most of the team not sleeping for the full 24 hour period. We are very proud of the finished product we managed to produce in this short amount of time. Managing to use the Clarifai API effectively was also very rewarding.
What we learned
What's next for eatr
At the moment, the web application is not designed to handle a large number of simultaneous users. We would like to add the ability for a single user to track their daily intake, and set goals for themselves. We all feel very passionate and proud of this project, and we hope to work on it further together.