A brief introduction to what we’ll do.

This guide helps you export your Teachable Machine model and let it interact with a Arduino Microcontroller.

Within this, we will use Teachable Machine to set up an train a simple ML model.

We’ll export this as a p5.js (or Processing) sketch and run this in our browser, connected to our webcam.

The Processing sketch will open Serial communications and relay what it’s seeing on the webcam (as classified by the TeachableMachine Model).

The Arduino, connected via USB to the laptop where the p5.js script is running, will receive the message on the serial port, and act accordingly.

Create your model in Teachable Machine

EXPORT THE MODEL

Choose Export Model and then you are presented with a variety of options. We’ll work with the TensorFlow.js which is designed to export the ML model to use in web browsers. We’ll insert some Javascript to make a request to the Particle Cloud and relay what our web browser based model is discovering to our Particle board.

• First choose the option to upload your model and copy the sharable link. You'll need this later.
• You can either choose to export Javascript or p5.js. We’ll use the p5.js option for this demo