"Octopet" is a m-health Android application focused on food logging. The idea was that users would take pictures of their food, and the application's resident "Octocat" would determine whether that food was healthy or not based on an algorithm involving natural language/image processing and food databases. Stemming from our love of pets and unhealthy junk food consumption, we wanted to create an application that promotes a healthy diet incentivized by sustaining the life of a virtual companion.
Octocat provides a fun and responsible way to track your dietary health! By snapping a picture of what you eat on the app, you will also feed your pet with it. Octocat will then respond depending on how healthy the food is. If unhealthy, Octocat will gag. If healthy, Octocat will beam! Be careful though, a cat only has nine lives and Octocat can't take too much punishment. ...We primarily utilized the Android Studio IDE to build Octocat. Using Java and XML files, we connected our app to the Android phone's camera for picture-taking capability. Then, we utilized image labeling from Google's Firebase machine learning kit and visual API to process and label images. Finally, we brought in our Octocat from Giphy via the Transposit API.
There is no demo, but our source code is on Github and our promo video is here!
More information: https://devpost.com/software/octopet