Di Pizza

Using our webapp you can find the most sustainable ingredients for your pizza. For this reason simply take a picture with your phone and see ingredients with their co2 emission and nutritional value.

Comment

Inspiration

Using computer vision we wanted to discover new ways to interact with products. We were looking for a specific use case that was doable in the short amount of time. We all loved pizza and from often checking out pictures of pizzas at restaurants, we thought we could detect ingredients and suggest the most sustainable product from the product assortment to buy it in your local grocery store to make your dream pizza! Obviously the idea would be to extend this to all kinds of food and not only specifically pizza.

What it does

One can easily take a picture or upload a picture of their favorite pizza. Then, we detect the used ingredients and find the most sustainable product from the product range in your local grocery shop.

How we built it

We built it using the API from Migros for product search, cross-referencing it with the API that IBM offered in order to get a CO2 footprint. The backend was built with python and the frontend as a webapp using javascript, html and css.

Challenges we ran into

Coordination accross timezones was not that easy, but in the end we think we successfully overcame this challenge! Furthermore, we had a lot of trouble getting an adaptable computer vision model set up in order to properly detect ingredients using state-of-the-art ML algorithms. In the end, we had to give up due to time constraints and went with plan B using a smaller neural network with faster training times. One team mate had trouble with energy supply in India, hence, we had to properly coordinate to not lose the made progress.

Accomplishments that we are proud of

We are proud to integrate a random team mate from around the globe. We are proud on the running prototype. We are proud on our organization. We are proud to be part of the largest hackathon in Europe!

What we learned

We learned to prioritize time planning such that it is possible to have an MVP asap and then keep expanding from there. We learned about MongoDb. Some of us learned about flask as a dev tool.

What's next for Di Pizza

Win hackzurich 2020!t

+ 17 more
Share this project:

Updates

Submission history