Teachable Machine: Build Image, Sound & Pose Detection Models

In this interactive session, students ages 7–10 (grades 3–5) learn basic AI concepts and build their own recognition models using Google’s Teachable Machine. The class emphasizes hands-on learning: kids collect examples, train models for image classification, sound detection, and pose recognition, then test results through live demos and short challenges. No coding experience is required. The workshop begins with a clear, kid-friendly explanation of how AI learns from data and why training examples matter. Students then follow guided steps to set up Teachable Machine, gather sample inputs (images, sounds, body poses), and train separate models. Each model is tested in real time so learners can see how the AI responds to different inputs and learn how to improve accuracy by adjusting examples. Throughout the session, the instructor demonstrates techniques, prompts small experiments, and leads brief challenges that encourage students to refine their models and think about where these technologies appear in everyday life—such as games, camera filters, and accessibility tools. The class keeps activities short and focused so every student gets practical experience training and testing at least one model type. This single-session format balances explanation, step-by-step guidance, and hands-on practice to give learners a clear, practical introduction to AI model building using a visual, no-code tool.