Machine learning with Google

 


Creating an ML model seems like a complex and time-consuming task to complete, but it doesn’t have to be this way. I will show you how to easily you can create and train an ML model using Teachable Machines

It is a Google Cloud’s no-code platform that trains ML models to recognize images, sounds, or poses. Let me show you with an easy example —

Project Aim: To train the computer to recognize 2 objects

Step- 1: Go to Teachable Machine Website

Step- 2: Select Image Project (As per our requirement)

Step- 3: Let's change the names of Class 1 & Class 2 with Book & Pen

Step- 4: Upload pictures of the objects, In our case “Book and Pen”

Tip: More pictures will better train the computer

Step- 5: Click on Train Model *After uploading

The results are Here!



As you can see, the book class is 100% and the pen class is 0%…

This means that the computer is 100% sure that I'm holding a book, not a pen.

If you didn't get the same result as mine, Try to take the training pictures in a well-lit room.

And take 40–50 training photos which will help the computer to better understand the difference.

Hope you like this article:)

— -My Project — —

Post a Comment

0 Comments