Food Track

A React Native (Expo) app for logging meals by photo. Take a picture of food, add a short description, and receive volume and nutrition estimates from the companion Food Volume Estimation and Food Natural Language Processing APIs.

Prerequisites

• Node.js and npm
• The Food-Volume-Estimation API running on a machine reachable from the device
• The Food-Natural-Language-Processing API running on a machine reachable from the device
• An Android/iOS device or simulator with camera access

Setup

npm install

Create a .env file with the API server URL:

EXPO_PUBLIC_API_URL=http://<your-ip>:8001

Then run:

npx expo start

For a physical device, use the IP address of the computer running the API, not localhost. The app submits images to /api/v1/submit and polls the job endpoint until the estimate is ready.

Checks

npx tsc --noEmit
npm run lint

Structure

• src/app/ - screens, file-based routing via expo-router (capture -> describe -> processing -> log detail/edit)
• src/components/ - UI components, grouped by screen/feature
• src/hooks/ - shared state (logs, theme, onboarding), persisted with AsyncStorage
• src/services/volumeEstimation.ts - submits and polls volume-estimation jobs
• src/services/logSubmission.ts - persists a pending job and maps its result back to a log
• src/types/ - shared TypeScript types (Log, FoodNutrients)

Meal logs and captured images are stored locally on the device using AsyncStorage and the app document directory. They are not uploaded anywhere other than the configured estimation API.