Author

Mateusz Matkowski

About

This is a React Native project for LEGO sets classification created for my Master Thesis. It is using a TensorFlow Lite model with the help of react-native-fast-tflite library. It can predict 50 different LEGO sets with the validation accuracy of 96%. After choosing an image from the galery (left button) and clicking the right button it returns up to 5 most similar sets to the input image.

Project bootstrapped using @react-native-community/cli.

Screenshots from the app

Empty page:

Model results:

Getting Started

Note: Make sure you have completed the React Native - Environment Setup instructions till "Creating a new application" step, before proceeding.

Step 1: Start the Metro Server

First, you will need to start Metro, the JavaScript bundler that ships with React Native.

To start Metro, run the following command from the root of your React Native project:

# using npm
npm start

# OR using Yarn
yarn start

Step 2: Start your Application

Let Metro Bundler run in its own terminal. Open a new terminal from the root of your React Native project. Run the following command to start your Android or iOS app:

For Android

# using npm
npm run android

# OR using Yarn
yarn android

For iOS

# using npm
npm run ios

# OR using Yarn
yarn ios

If everything is set up correctly, you should see your new app running in your Android Emulator or iOS Simulator shortly provided you have set up your emulator/simulator correctly.

This is one way to run your app — you can also run it directly from within Android Studio and Xcode respectively.