Inspiration: It's time that everyone became less confused about how to eat right. We thought this was the perfect opportunity to explore solutions.
What it does: This iOS app automates meal recommendations based on the the user's inputted characteristics, medical condition, and goals. Many of us face decision fatigue when we're confused about how to use leftover ingredients, how to make a healthy meal out of it, and what to make for the next meal. Our object-detection feature facilitates these decisions by recommending recipes that will fit the user's nutritional needs, goals, budget, and preferences. The user has the option to track a serving of the recipe they've made which will be analyzed by an existing API (on Edamam-Nutrition) so that users can get immediate feedback about how they can reach their nutrition targets for the day (with suggestions: "have you tried this recipe?", "this recipe may help you achieve your protein goals for the day", etc). User's ultimately gain time by spending less time deliberating and more time cooking and eating well; the app simultaneously tackles food waste by encouraging the use of leftovers.
How I built it: We built a mobile application through SwiftUI, and used ARKit2 for object-detection (of food). For our own learning experience, we attempted to build the training model from scratch and spent hours collecting data to detect enough detail on different fruits. Once the food was detected, users are given an option to find recipes containing those detected foods which was enabled with the use of Edamam-API. The nutrition facts were extracted from those recipes from the same API to be compared with the the individual's nutrient requirements.
Challenges I ran into: 3-D detecting the food was a great hurdle to say the least. Learning how to update every part of my macOS system and downloading packages for more than a few hours was a learning curve as well for the new hacker in the team.
Accomplishments that I'm proud of: We're proud about working with complete strangers for the first time, being extremely resourceful throughout, learning completely novel technologies, having fun, and overcoming the hottest sauce alive. The experienced were very patient to the newbies and the newb was willing to learn.
What I learned: We're proud that we know how to navigate ARKit2 and Swift, which were two technologies that we are grateful to learn.
What's next for Smhacked: Smhacked was a medium for a group of passionate, perserving, and patient hackers to unite and explore a completely new technology. Each individual tackled an unfamiliar territory, and these skills will transcend Smhacked into future endeavors because of the people we met and the programs we risked to explore. Smhacked was an adventure that may see the day again with sudden inspiration or with greater practice with the programs we used.