Presented by:

Ken Kahn

from University of Oxford

Teachable Machine has a very simple interface for creating machine learning models that can provide labels for images, sounds, and poses. I'll demonstrate a new Snap! block that imports these models for use in your own projects.

Using this block, you can create projects that classify images, respond to different sounds, and much more. You can train Teachable Machine to recognise different hand gestures or facial expressions and then make games that respond to hand movements or different faces. Or, you can make a project that can identify a bird (or animal or insect) by its appearance or sound.

All this works on your computer and no images or sounds ever leave your device. This ensures privacy. Alternatives that contact servers for image or sound recognition also introduce a delay due to communication that can make it too slow for use in a game or other app.

An illustration of the use of this block can be found in this project.

Learn how to create your own AI project in Snap!

Duration:
5 min
Room:
Plenary
Conference:
Snap!Con 2021
Type:
Lightning Talk
Track:
Lightning Talks