Groceries with machine learning

This project extends the image classification tutorial with something more practical: a robot that does the shopping for you!

If you haven’t followed the introduction to machine learning yet, be sure to check it out first. In this project, the robot is trained to recognize various categories of groceries, such as strawberries, broccoli, or cake.

When you try this project, be sure to train your own model with pictures of your own items (toy groceries or otherwise), otherwise the robot won’t recognize anything.

This program makes the robot drive past groceries spaced 300mm apart, checking each item against the broccoli category and driving to it when found.

You can see the result in the video below.