AI Calorie Tracker

Inspiration

I found nutrition tracking incredibly difficult while eating at dining halls during college. I tried other apps (MyFitnessPal) and they took super long to track foods which was really tedious. With vision LLMs coming out, we can instantly track average portion average sizes of food that we eat to get a rough estimate of calories without the tedious manual work.

What it does

The app simplifies nutrition tracking through visual food logging. You simply have to take a photo of your photo to track calories and macros. We also keep all the photos of the food you ate!

How we built it

React Native + Serverless functions. Deployed using Expo.

Challenges we ran into

Storing photos and running them for quick AI inference especially when the user navigates off the app.

Accomplishments that we're proud of

Building the AI vision model. We use image segmentation + a specific prompt thats grounded to the USDA database.

What we learned

How to build an app in React Native. Design concepts in Figma that change the feel of the app.

What's next for AI Calorie Tracker

We're trying to advertise our product more and see how our users like it. We also want to build a gallary of foods for people to look back on all their meals and analytics to see your fitness journey.

Built With

• native
• react