Google Teachable Machine

What is Google Teachable Machine?

Google Teachable Machine is a tool that lets you train a computer to recognize your own images, sounds, and poses — no coding required. The platform makes it simple to create custom machine learning models you can use in websites, apps, or other digital content. Examples include adding image recognition to an e-commerce site, building voice-controlled apps, or creating gesture-based interactions. You upload your data, and the platform’s machine learning handles the rest, removing the need for coding or data science knowledge.

Who created Google Teachable Machine?

Steve Saling developed Google Teachable Machine. He originally built it to explore training systems for non-standard speech interactions, specifically for people with ALS, stroke survivors, and others who need extra help communicating. He has also used the tool in Project Euphonia, finding ways for people to communicate through facial gestures that trigger sounds.

What is Google Teachable Machine used for?

• Build apps controlled by voice.
• Add image recognition to e-commerce sites.
• Create custom interactions that respond to gestures.
• Train models to classify images, sounds, and body positions.
• Build custom machine learning models for websites and apps.
• Explore machine learning concepts like classification and how technology impacts society.

Who is Google Teachable Machine for?

• Web developers looking to add smart features.
• App developers creating more interactive experiences.
• Digital content creators aiming to engage their audience.
• E-commerce website developers seeking to improve functionality and user experience.

How to use Google Teachable Machine?

1. Visit the Teachable Machine platform on Google’s website.
2. Upload your own data — images, sounds, or videos you want the computer to learn from.
3. The platform’s machine learning algorithms analyze and learn from your data.
4. Train and refine the system by giving it feedback and corrections to develop accurate models.

The platform handles the complex coding, so you can focus on building models tailored to your needs and implementing them wherever you need them.