How to Teach Computer Vision to Families

A photograph reaches a computer as a grid of numbers, one per pixel, describing colour and brightness. Computer vision is the work of turning that grid into something useful: a label, a box around an object, a count, a measurement. Early layers of the system detect very simple things — an edge here, a change in brightness there. Later layers combine those into shapes, then into parts, then into objects. Nothing in the process involves the computer seeing in any human sense; it is arithmetic on a grid of numbers, repeated at enormous scale, tuned by examples until the output tends to match what a person would have said.

The part worth getting right early is the misconception. Most people assume that a camera plus software means the computer "sees" the room. In fact it processes one frame of numbers at a time with no memory of the room, no sense of depth unless explicitly given it, and no idea that the objects it labels continue to exist when the frame changes. Correcting that once, early, saves a great deal of confusion later — and it is the single idea most likely to stick with the children you teach.

Reading level: Mixed — the adult reads, the children do. Maths assumed: Whatever the youngest participant can follow. Realistic focus in one sitting: about 30 minutes. Pushing past that produces activity, not learning — the child keeps clicking but stops forming a model of what is happening.

Supervision: The adult is a participant rather than a supervisor. Half an hour together beats three hours apart. The value is in the conversation the activity provokes.

The mixed-age problem is real and has a clean solution: do not scale the activity down to the youngest, split the roles instead. In a classifier activity the youngest child collects and sorts the examples, the middle child runs the training, and the oldest designs the test that tries to break it. Everyone is working on the same artefact at their own ceiling, and nobody is watching.

The payoff of doing this as a family rather than individually is the disagreement. When a nine-year-old and a fifteen-year-old predict different results and then watch what actually happens, the conversation afterwards does more than the activity did. Build in the prediction step explicitly — ask everyone to commit to a guess out loud before you run it.

Become the camera (about 10 minutes, ages 3–7). You need: A cardboard tube or rolled paper. 1. Have the child look at the room through the tube and describe only what fits in the circle. 2. Move the tube and ask what happened to the thing they just described. 3. Ask whether the sofa stopped existing when it left the circle. You will know it worked when the child can explain that the camera only knows what is inside the frame right now.

Pixel grid on graph paper (about 20 minutes, ages 5–10). You need: Graph paper and two coloured pencils. 1. Fill in squares to draw a simple shape — a heart or a letter. 2. Read the grid out loud row by row as "filled, empty, filled". 3. Have a second person redraw the shape from the read-out alone. You will know it worked when the child can explain that a picture can be sent as a list of numbers and rebuilt exactly.

Break a classifier on purpose (about 35 minutes, ages 8–15). You need: A laptop with a webcam, Teachable Machine. 1. Train a classifier to tell two of the child's toys apart. 2. Test it in a different room, under different light, at a different distance. 3. Log every condition that caused a wrong answer. 4. Retrain covering those conditions and re-measure. You will know it worked when the child can name at least two conditions that change the answer without changing the object.

Audit a real vision system (about 45 minutes, ages 12–18). You need: A phone with Google Lens or a similar app. 1. Pick twenty household objects and predict which the app will get wrong. 2. Test all twenty and record the actual result against the prediction. 3. Group the failures — was it lighting, angle, an unusual object, or an object type the app was never built for? 4. Write two sentences on who would be harmed if this system were used for something that mattered. You will know it worked when the teenager can distinguish a failure caused by input conditions from one caused by training coverage.

The shortlist is deliberately short. A child who uses one tool properly and finds its limits learns more than one who samples six. Start at the top of this list and only move on when the current tool stops being able to answer the next question.

The most common failure is starting with the mechanism instead of the behaviour. Adults reach for how the system works internally, because that is the interesting part to an adult. A child needs to see the thing behave — make a right guess, then a wrong one — before any explanation of the internals means anything.

The second failure is treating a correct output as the end of the lesson. The learning is concentrated in the failures: the lighting that broke the classifier, the accent it could not parse, the example nobody thought to include. Budget deliberate time for breaking the thing on purpose, and treat every break as the result rather than as a problem to hide.

The third is over-supervising or under-supervising relative to age. The adult is a participant rather than a supervisor. Getting this wrong in either direction costs you — too little and the session drifts, too much and the learner stops making the guesses that teach them anything.

First, every age given is the age the tool becomes genuinely usable, not the vendor's marketing age. Those differ often. Nothing in this collection was excluded on age grounds for this band.

Second, only tools with a genuinely free tier are listed — free meaning a real project can be finished without paying, not a trial that expires mid-activity. 4 of the 5 can be used with no account at all: Google Teachable Machine (image project), Quick, Draw!, Scratch with the video-sensing extension, OpenCV with Python. That matters more than it sounds at this age, because an account is a data-collection decision a parent has to make on a child's behalf.

Third, "no screen tool is appropriate yet" is treated as a valid answer rather than a gap to fill. Where this page recommends physical objects over software, that is the recommendation, not an omission.

You can verify all of this yourself in about ten minutes: open each tool listed, check whether it demands an account or payment before producing anything, and see whether the child you have in mind can reach a first result without an adult reading the interface aloud. If any recommendation here fails that test, it is wrong and worth telling us about.