Like taking pictures of your food? Have Food.C track what you eat to help you eat healthier. We use Keras to detect the content of the photos you take, then match it against its nutrition facts, so you can eat healthier while instagramming your meal. We even use facial recognition to track user profiles and nutritional intake.
How we built it
We run a flask server with a web application, then we write the camera data to the server on a photograph. This is then classified by Keras (VGG16), and then checked against a SQLite database built from the government nutritional data.
Challenges we ran into
We ran into some strange encoding bugs when we were sending the images to the classifier. Additionally, using the ImageNet model, and a weak wifi connection made it harder to research things so we had to get resourceful.
Accomplishments that we're proud of
We didn't have a clue what we were making until late into the hackathon! But we crunched and got it done.
What we learned
Come prepared! But don't worry about changing plans. And eat your veggies!
What's next for Food.c
Taking over the world!