Teachable Machine is a web utility that is a quick, simple way of making machine learning models for a project without any prior knowledge of programming. It can train based on images, sounds, or poses that the user uploads to the site. The model can be exported to any site or app by using the JavaScript or p5.js (a JS client-side library) code provided when the model is fully trained.

I made three web apps using Teachable Machine and it was a seamless experience. Once I made the first web app by using p5.js, it was very easy to make another one by replacing the model link of the first web app with the link of the new model.

Let’s say you want to build a facial detection machine learning model using Teachable Machine (like I did). The first step is click the link below and click on “Image Project”. Rename the two classes to something like “Face” and “NoFace”. Record many pictures of your face and then do the same without your face in the frame of the pictures. After that, click on “Train Model”. When that’s done, click on “Export Model” and then on “Upload my model”.