Wolfram’s Artificial Intelligence Can Identify Objects In Your Photos
SHARE

The Wolfram Language Image Identification Project showcases just one function out of the 5,000 that are built into the Language. It can automatically identify the subject of any image you choose. Wolfram Research is the same company that created the science behind Apple’s automated assistant, Siri.

ImageIdentify can be used on your desktop or phone, and we found the results to be pretty accurate. You simply drag or upload an image, and the program will instantly identify it, and give you an answer.

httpswww.popphoto.comsitespopphoto.comfilesscreen_shot_2015-05-14_at_2.18.34_pm.png

After ImageIdentify has named what is in your image, it asks you to rate how it did. The ratings range from “Great!” to “What the heck?!” If you rate anything less than Great, the program will ask you what you think would be a better label.

ImageIdentify is a function that relies on complex algorithms and built-in knowledge, but manifests as a simple and easy-to-use site. The tool uses machine learning in order to become more accurate over time. The more images the AI sees of one subject, the more precisely it can identify it in the future.

Wolfram’s AI uses a neural network much like our own brain does, comprised of multiple layers in order to process details. Each layer filters a different amount of detail, starting with the broadest aspects and narrowing to the subtlest.

“It won’t always get it right, but most of the time I think it does remarkably well,” writes founder Steven Wolfram in his blog post, “And to me what’s particularly fascinating is that when it does get something wrong, the mistakes it makes mostly seem remarkably human.”

The ImageIdentify function will be made accessible to developers, so the language could be used in APIs and apps in the future.