The Rubbish Sorter
Build an image recogniser from start to finish - using everything you've learned.
What you'll learn in this level
- How the steps of an AI project fit together
- Why every choice shapes the next one
- What people have to decide along the way
Your first real AI project
A school wants a bin that recognises by itself whether something is paper, plastic or food waste. You're the AI developer.
You'll build the system in five steps. At each step you have to make a real decision - and each one comes from a different module you've already worked through.
How it works
- 1
Choose the groups
You decide which kinds of rubbish the system should tell apart.
- 2
Collect pictures
You collect pictures for each group. The more different they are, the better.
- 3
Label
For every picture you write down which group it belongs to.
- 4
Test and improve
With new pictures you check where the system still gets it wrong.
Guess first, then experiment
Your rubbish robot should get better. What helps most?
Your first real AI project
A school wants a bin that recognises by itself whether something is paper, plastic or food waste. You're the AI developer.
You'll build the system in five steps. At each step you have to make a real decision - and each one comes from a different module you've already worked through.
What you need from three modules
- • Machine Learning: which training data? How do you test whether it really works?
- • Neural Networks: how big does the network need to be?
- • Computer Vision: labelling images and knowing the limits of recognition.