Build a Teachable Machine Image Classifier With Your Kid

Introduce your child to the mechanics of machine learning by training a custom computer vision model in fifteen minutes.

Most children interact with artificial intelligence daily, whether through video game matchmaking, video recommendations, or voice assistants. However, they rarely get to see behind the curtain to understand how these systems actually learn. By building a custom image classifier, you can demystify AI, transforming your child from a passive consumer of technology into an active builder. This project requires no coding experience, no paid subscriptions, and no specialized hardware (just a computer with a standard webcam). It offers a hands-on, highly engaging way to build true technological intuition.

The Tool: Google's Teachable Machine

To build our image classifier, we will use a free tool developed by Google called Teachable Machine. It runs entirely in your web browser, meaning no data is sent to external servers, and no account creation is required. This makes it an exceptionally safe and accessible environment for children aged eight to sixteen.

To get started, open a web browser on your computer and navigate directly to https://teachablemachine.withgoogle.com/ . Click on "Get Started" and select "Image Project," then choose the "Standard Image Model."

Step-by-Step Project Guide

We recommend building a classifier that can recognize three different hand gestures or three different household objects (such as a mug, a pen, or a notebook). For this guide, we will build a "Rock, Paper, Scissors" hand gesture detector.

Step 1: Define Your Classes

In machine learning, a "class" is a category of data. On the screen, you will see two default classes labeled "Class 1" and "Class 2." Rename "Class 1" to "Rock" and "Class 2" to "Paper." Click the "Add a class" button at the bottom to create a third class, and name it "Scissors."

Step 2: Gather Training Samples

Now, click the "Webcam" button under the "Rock" class. Have your child hold their hand in a fist (the "Rock" gesture) in front of the camera. Click and hold the "Hold to Record" button.

Instruct your child to slowly rotate their hand, move it closer to the camera, and move it further away while you record. This ensures the model learns what a fist looks like from multiple angles. Aim to capture about one hundred image samples. Repeat this exact process for the "Paper" class (flat hand) and the "Scissors" class (two fingers extended).

Step 3: Train the Model

Once all three classes have at least one hundred samples, click the blue "Train Model" button in the middle column. It is crucial to leave the browser tab open and untouched during this process. The computer is now analyzing the pixels in the images, finding patterns that distinguish a fist from a flat hand or extended fingers. This usually takes about thirty seconds.

Step 4: Test and Break the Model

Once training is complete, the "Preview" panel on the right will activate your webcam. Have your child make the different gestures and watch the output bars at the bottom. The bars will show, in real time, the percentage of confidence the model has in its prediction.

Now, try to "break" the model. What happens if you use your left hand instead of your right hand? What happens if you change the lighting in the room, or if a parent steps into the background? This is a fantastic opportunity to show how models can become confused when presented with data they have not seen before.

Parent & Educator Sidebar

Core AI Concept: Supervised Learning This activity demonstrates supervised learning, where an algorithm learns to associate inputs (images) with labels (classes) provided by a human. The model does not actually "know" what a hand is. It simply recognizes mathematical patterns in the pixels.

Conversation Starters:

• "Why do you think the model got confused when we turned off the lights or used a different hand?" (Discusses the importance of diverse training data).
• "How do you think self-driving cars use this technology when they look at stop signs or pedestrians?" (Connects computer vision to real-world safety).
• "If we only trained the model with my hand, would it work well for your hand?" (Introduces the concept of algorithmic bias).

Action Steps Summary:

• Open the Tool: Go to https://teachablemachine.withgoogle.com/ on a laptop or desktop computer.
• Create Classes: Set up three distinct categories (such as Rock, Paper, and Scissors).
• Record Samples: Use your webcam to capture one hundred varied images for each category.
• Train and Test: Click "Train Model" and then test its accuracy using real-time webcam input.
• Discuss the Tech: Use the sidebar questions to help your child understand how this relates to the AI systems they use every day.

Share your child's custom classifier with family members and start exploring the mechanics of machine learning together.