Build a Teachable Machine Image Classifier With Your Kid

Demystify artificial intelligence by building a real-time computer vision model using a standard webcam.

Artificial intelligence is no longer a futuristic concept: it is the defining technology of our era. For children aged 8 to 16, understanding how AI works is just as important as learning to read or write. However, passive consumption of technology does not foster true comprehension. To demystify these complex systems, children must actively build them. Google's Teachable Machine is a free, web-based tool that allows anyone to train a machine learning model using a standard computer webcam. This hands-on project guides you and your child through building an image classifier, providing a practical introduction to the fundamentals of computer vision and neural networks.

Demystifying AI for the Next Generation

Many children interact with AI daily through voice assistants, video game algorithms, and social media feeds. Despite this familiarity, the underlying technology remains a black box. This lack of understanding can lead to either overestimating AI's capabilities or fearing its influence. By building a functional image classifier, children learn that AI is not magic: it is a system of patterns, data, and mathematical rules. This project requires no coding experience, making it accessible for younger children while offering deep conceptual insights for teenagers.

Setting Up Your Workspace

To begin, open a web browser on a laptop or desktop computer equipped with a webcam. Navigate directly to the Google Teachable Machine platform at https://teachablemachine.withgoogle.com/ . Click on the Get Started button on the homepage. You will be presented with three project options: Image Project, Audio Project, and Pose Project. Select Image Project , and then choose Standard Image Model . This opens the main workspace, which is divided into three clear sections: Training Data, Training, and Preview.

Step 1: Defining Your Classes

In machine learning, a class is a category of data. Your model will learn to sort incoming visual information into these classes. For this project, select two distinct objects from around your home. Excellent choices include a coffee mug and a pen, a green apple and a yellow banana, or even two different hand gestures like a thumbs-up and a thumbs-down. In the first box labeled Class 1, click the pencil icon and rename it to match your first object. Rename Class 2 to match your second object. This step teaches children the importance of labeling data accurately.

Step 2: Gathering Training Data

Now, click the Webcam button under your first class. Hold your first object in front of the camera and press the Hold to Record button. As you record, slowly rotate the object, move it closer to and further from the lens, and tilt it in different directions. Capture at least one hundred images to provide the model with a diverse dataset. Repeat this exact process for the second class using your second object. Explain to your child that the computer does not see objects the way humans do: it analyzes pixels, colors, and edges. Providing varied angles helps the computer learn the unique features of each object.

Step 3: Training the Model

Once you have gathered sufficient images for both classes, click the Train Model button in the middle column. This process takes only a few seconds as the browser-based neural network analyzes the images to find patterns. During this phase, the computer associates the visual patterns of the first object with its label, and the patterns of the second object with its label. This is known as supervised learning. Instruct your child to watch the progress bar as the model trains, explaining that the computer is adjusting thousands of internal parameters to minimize errors and improve its accuracy.

Step 4: Testing and Iterating

After training is complete, the Preview panel will activate, showing a live feed from your webcam. Hold up one of the objects and watch the output bars at the bottom of the screen. The model will display a percentage representing its confidence in identifying the object. To make this an active learning experience, challenge your child to try and trick the model. What happens if you hold the mug upside down? What happens if you show it a completely different object, like a book? If the model makes a mistake, discuss why it occurred and how to fix it by adding more diverse training images to the classes.

Parent and Educator Guide

Core AI Concept: Supervised Learning Supervised learning is a type of machine learning where a model is trained on labeled data. The computer is given inputs (the images) and the correct answers (the labels). By analyzing thousands of examples, the system learns to identify patterns and make predictions on new, unseen data.

Conversation Starters:

• How does the computer know the difference between the mug and the pen if it has never seen them before today?
• What would happen if we only trained the model with red apples, and then showed it a green apple?
• How could a self-driving car use this exact technology to keep pedestrians safe on the road?

Action Steps Summary

• Access the Tool: Open a web browser and go to https://teachablemachine.withgoogle.com/ .
• Select Your Objects: Choose two distinct household items and label Class 1 and Class 2 accordingly.
• Record Training Data: Capture at least one hundred webcam images of each object from various angles and distances.
• Train the Model: Click Train Model and observe the browser-based neural network process the data.
• Test and Debug: Use the live preview to test the model, identify errors, and add more training data to improve accuracy.

Want more engaging, hands-on projects to explore technology with your family?