Inspiration

Machine learning and image recognition was a particular technology that members of our group had discussed about intensively. It stemmed from a simpler hack involving superpositioning a banana over every detected faces in a photo.

What it does

Determine and track the calories intake of the user.

How we built it

While each of our members was responsible for specific aspects of the hack, we constantly hover each other to collaborate on features and work out kinks in the code.

The project was sketched out on the back side of ads found all over the sdhack area. The app was developed and baked in four laptops running Android Studio (linked by Git). Firebase was chosen to host our (and user) data online. Custom assets were hastily drawn and put together in GIMP and Inkscape.

Challenges we ran into

Our main focus was to integrate both CLARIFAI's API and SNAPKIT's API into banana™. However, SNAPKIT in particular was throwing extra-dimensional errors that we were humbly humbled by and were ultimately defeated by.

Accomplishments that we're proud of

We piled our miniscule experiences and diverse technical skills into an awkward but functional driving force that ended up creating a project that tested and pushed our capabilities.

What we learned

E V E R Y T H I N G

What's next for bananas

This isn't even bananas™'s final form. We intend to continue with banana™ with several addtional features and improvements: - integrate banana™ into UCSD's dining hall experience - integrate banana™ into our other startup PLATES business - take advantage of CLARIFAI's machine learning API and allow users to contribute to banana™'s image recognition - clean up the UX/UI to sooth ours (and the clients') aesthetic needs - refactor the code to make the app more responsive - stupid-proof banana™

Built With

Share this project:

Updates