AI Projects with Teachable Machine & Other Machine Learning Models

The integration of technology in education is transforming the way students learn and engage with complex concepts. Machine learning, a subset of artificial intelligence, is becoming increasingly accessible and relevant in classrooms, offering innovative ways to enhance learning experiences. One of the most user-friendly tools available is Teachable Machine, which allows students to create machine learning models without needing extensive programming knowledge. We will explore various school projects that leverage Teachable Machine and other machine learning models, providing educators and students with practical ideas and insights to incorporate these technologies into the curriculum to foster creativity and critical thinking skills in the classroom.

What is Machine Learning?

Machine Learning (ML) is like giving computers a brain that learns from data. Instead of programming a machine to perform specific tasks, we feed it a bunch of information, and it figures out how to handle problems on its own.

Importance of Machine Learning in Modern Education

In today’s education system, AI can personalize educational experiences, help identify students’ strengths and weaknesses, and can even automate administrative tasks. Introducing students to machine learning empowers them to understand the engine driving this trend, and how to use it to do exactly what they want it to do. This will prepare them for careers in a world where AI is prevalent.

What is Teachable Machine?

Teachable Machine is a web-based tool that makes creating machine learning models fast, easy, and accessible to everyone. You train a computer to recognize your images, sounds, and poses without writing any machine learning code. Then, use your model in your own projects, sites, apps, and more. Perfect for school projects, it takes the guesswork out of machine learning, allowing students to focus more on creativity and less on code.

These models use a technique called transfer learning. There’s a pretrained neural network, and when you create your own classes, you can sort of picture that your classes are becoming the last layer or step of the neural net. Specifically, both the image and pose models are learning off of pretrained mobilenet models, and the sound model is built on Speech Command Recognizer. You can check out all of the machine learning code powering this app at the Teachable Machine Communities repo on github.

Benefits of Using Machine Learning Models in School Projects

Active learning, a pedagogical approach that emphasizes student engagement and participation, is being revolutionized through the use of interactive technologies such as machine learning models. By combining hands-on experiences with tech tools, students are engaging with their learning like never before.

Active Learning with Technology in Elementary Education is all about getting kids off their seats and actively involved in the learning process. Rather than just about absorbing information passively students are actually doing, creating, and exploring.

Engagement and Interactivity

Machine learning tools bring projects to life by engaging students in hands-on, interactive experiences. With the ability to create their own models, students become active participants in their learning journey rather than passive listeners. Whether they’re training a model to recognize their dance moves or sorting fruit by color, the sense of ownership leads to increased enthusiasm and excitement for learning.

Hands-On Learning Experiences

Teachable Machine and similar tools provide an active approach to grasping complex concepts in machine learning. Students can explore through trial and error, experiment with different datasets, and witness firsthand how their models perform. This experiential learning is not only fun but also reinforces critical thinking and problem-solving skills.

Getting Started: Setting Up Your Teachable Machine Lesson

Required Tools and Resources

To get started with Teachable Machine, all you need is a computer or laptop with an internet connection and a modern web browser. It’s also helpful to have a camera (if you’re using the image model) and some creativity to fill the project with lots of exciting data. Then use Google’s Teachable Machine website to access the platform and other useful resources.

Step-by-Step Guide

Getting up and running with Teachable Machine is simple. Just follow these steps:

1. **Open your web browser** and navigate to https://teachablemachine.withgoogle.com/.
2. **Choose your project type** (image, audio, or pose) based on your interests.
3. **Gather your data**—take pictures, record sounds, or strike a pose, depending on your selected project type.
4. **Train your model** by clicking the “Train Model” button; sit back and watch as your data transforms into an intelligent creation!
5. **Test your model** to see how well it performs. If it’s not quite right, don’t worry—tweaking is part of the process!
6. **Export your model** to share your genius with the world.

And voilà! You’re all set to showcase your machine learning masterpiece.

Project Ideas for Machine Learning Models

Sold on the idea, but not sure where to start? Check out these project ideas!

Gesture/Pose Recognition Projects

A classic Teachable Machine project is creating a gesture recognition game. This project uses pose detection to recognize different body movements and translate them into game actions, like jumping over obstacles, dodging enemies, or collecting points.
Use Teachable Machine to recognize different gestures, such as raising your hands, twirling, or performing a Fist-and-Palm Salute. Then assign each gesture to trigger a different song or sound effect, creating an interactive music experience.
By combining pose detection and AI, you can create a virtual pet that can recognize your movements then respond and perform actions like waving back, jumping, or sitting.
Teachable Machine can be used to create a custom avatar that reacts to the player’s pose and movements. The avatar can be set to change its clothing, expression, or background based on the gesture performed.
Build a fitness tracker that uses pose recognition to identify exercises and track performance. The app can provide real-time feedback, such as displaying how many reps were completed or offering tips to improve posture and form.
With pose and facial expression recognition, you can create an interactive storytelling tool where characters react based on the player’s expressions. Train your Teachable Machine to recognize different emotions like happy, sad, angry, and surprised through pose detection or face tracking. Then make a story in which characters can change their behavior or dialogue based on the detected expressions.

Sound Activated Projects

BYOTM (Bring Your Own Teachable Machine) has students send text messages to family and friends using their personalized Teachable Machine speech recognizer. Map your trigger words to customized messages and send them to any phone number.
Sound-activated art combines audio recognition with visual effects. This type of project helps you generate a piece of art or animate a drawing in response to different sounds, like claps, snaps, or speech.
Using sound recognition, you can also create a voice-controlled game where player character actions are triggered by voice commands, such as “Fight”, “Dig”, or “Carry”.

Data Analysis/Sorting Projects

Bring out your Arduino to create this Tiny Sorter by Google Creative Lab.
Utilize Teachable Machine to analyze data from your experiments. For instance, if you’re studying plant growth, train a model to recognize healthy plants vs. wilting ones based on images. You’ll not only learn about machine learning but also gather valuable insights into your scientific inquiries.

The Awesome Teachable Machine List has a truly amazing selection and organization of even more projects if this lesson type becomes a favorite amongst your students.

Other Lessons & Programs Promoting Active Learning with AI

Google Arts and Culture has Odd One Out to practice identifying AI and Say What You See to practice prompting.
Day of AI brings free and open-source AI tools, curriculum, assessments, and teacher professional development materials developed by MIT RAISE to students and educators everywhere.
Snorkl supports conceptual mastery–providing instant AI feedback on recorded explanations, with feedback offered on students’ verbal and visual explanations.
For web design students, after you’ve reviewed their HTML code together and they still can’t find the issue, have them use AI to troubleshoot. Students will then complete a short reflection explaining what the bug was, how it was fixed & how can they avoid this bug in the future. Suddenly AI becomes turns into a tool for learning, not just a quick fix or a crutch.
For your students learning Python, with Codeacademy’s Build a Machine Learning Model skill path they will learn to build machine learning models using regression, classification, and clustering. Along the way, they will create real-world projects to demonstrate your new skills, from basic models all the way to neural networks.
Engage in critical thinking with a Raising Them Right type of project. Have students create a “polluted” build vs a more “pure” build then feed them both the same prompts to contrast the differences in the AI’s “decisions” based on how the tool was trained.

Tips for Successfully Implementing Machine Learning in Classrooms

Best Practices for Educators

First things first: don’t panic! Machine learning might sound like rocket science, but breaking it down into bite-sized pieces can make it digestible. Start by introducing basic concepts and progressively build up to more complex ideas. Use visuals, interactive demos, or relatable analogies to engage students. Encourage questions, and don’t be afraid to admit when you don’t know the answer—it’s a great learning opportunity for everyone involved. Finally, keep your projects relevant to students’ lives to ignite their interest and imagination.

If you find yourself wanting to dive in and become more of a subject matter expert, check out Stanford’s free course to attain a Machine Learning Specialization for Beginners.

Encouraging Student Collaboration and Creativity

Two heads are better than one, especially when it comes to tackling tricky machine learning concepts! Foster a collaborative environment by assigning group projects that require different skill sets—from coding to presenting findings. Encouraging students to brainstorm solutions together will not only enhance their understanding but also add a splash of fun. Integrate creativity by allowing students to choose their project topics. Whether it’s predicting the outcome of a new video game or creating a model to analyze their favorite music, the possibilities are endless. Remember, the goal is to have them think outside the box… or, in this case, the neural net!

Future Trends in Machine Learning for Education

Emerging Technologies in the Classroom

As we look ahead, expect to see classroom assistants powered by AI that can help with grading or providing personalized feedback. Virtual and augmented reality, combined with machine learning, will continue to enable immersive learning experiences where students can interact with historical events or scientific phenomena.

Preparing Students for a Machine Learning Future

As machine learning becomes more integrated into various industries, preparing students for this future is crucial. Schools should focus on building a strong foundation in computational thinking and data literacy, starting in the early grades. Encourage field trips to tech companies, internships, or workshops that let students engage with real-world applications of machine learning. Additionally, integrating interdisciplinary projects that combine skills from math, science, and the arts will help students become well-rounded, innovative thinkers. After all, the goal is to equip the next generation to not only understand machine learning but to master it—and maybe one day program it to improve the world in ways we haven’t even thought of yet!

As we continue to embrace the possibilities of machine learning in education, tools like Teachable Machine empower students to explore their creativity while developing essential skills for the future. By engaging in hands-on projects and collaborative learning experiences, learners can gain a deeper understanding of complex concepts and apply them in meaningful ways. We encourage educators and students alike to experiment with these technologies, as the journey of discovery and innovation is just beginning. With the right resources and inspiration, the potential for impactful projects is limitless.