What is Teachable Machine?

Teachable Machine is a web-based tool that makes creating machine learning models fast, easy, and accessible to everyone. You train a computer to recognize your images, sounds and poses without writing any machine learning code. Then, use your model in your own projects, sites, apps, and more. Anyone can use this like Educators, artists, students, innovators, makers of all kinds — really, anyone who has an idea they want to explore.

How do I use it?

1) Gather Data:

You can currently train Teachable Machine with images (pulled from your webcam or image files), sounds (in one-second snippets from your mic), and poses (where the computer guesses the position of your arms, legs, etc from an image). More types of training may be coming soon :)

2) Train Model:

Train your model by just clicking a single button no need for any smoothing or pre-processing required, Teachable Machine will train a model based on the examples you provided. All the training happens in your browser, so everything stays on your computer. then instantly test it out to see whether it can correctly classify new examples like poses, voice, or images as per prior input which used for training.

3) Export Model:

There are a few ways to save your work. You can:

Save your entire project to Google Drive. You can save your project to your Google Drive. This saves a .zip file that contains all the samples in each of your classes to Drive. You can then open that .zip again from Teachable Machine later to pick up where you left off.
Download your samples. You can download all the samples in each class, and upload them later if you want to keep working with the same data.
Download your model. If you download your model and close your tab, you can use that downloaded model later — and nothing is saved on any servers.
Save nothing. If you close your tab and do none of the above, nothing is saved in your browser or on any servers.

How does it work on the back?

Teachable Machine uses TensorFlow.js, a library for machine learning in Javascript, to train and run the models you make in your web browser.

These models use a technique called transfer learning. There’s a pre-trained neural network, and when you create your own classes, you can sort of picture that your classes are becoming the last layer or step of the neural net. Specifically, both the image and pose models are learning off of pre-trained mobile net models, and the sound model is built on Speech Command Recognizer.

#teachable-machine #adafruit #machine-learning #artificial-intelligence #arduino