From Pixels to Predictions

This is a free, ad-free, visual course about one question: how does a machine actually see an image? It starts with a single pixel and builds, animation by animation, all the way up to a convolutional network that can recognise what it's looking at.

Every idea is something you scrub through and watch happen, not just read about. There are 12 short chapters across four parts — from “an image is just numbers” to how a network learns. Formulas and code are tucked behind optional reveals, so you can stay in the pictures or dig into the details.

It's built for the curious — students, self-learners, and anyone who wants a real intuition for the foundations of computer vision in under an hour.