Stephen Wolfram, the brains behind the amazing Wolfram Alpha computational knowledge engine, has created another tool that could have a wide array of uses. Known as the Image Identification Project, Wolfram's new website can identify what's in an image in just a few seconds.
Image Identify does a surprisingly good job of identifying the contents of an image, especially if it's a common object. The tool can even distinguish between types of animals, correctly identifying the difference between bactrian and dromedary camels, or golden retrievers and labrador dogs, in our brief testing.
Like many other tools that analyze images and try to recognize their contents, Google Goggles being one example, Image Identify doesn't always get things right. Wolfram says that the tool has trouble identifying art, specific people, and things that aren't everyday objects, while noting that it learns more each time it analyses an image and receives feedback from a human.
While it's great to have a website that can identify images, the inclusion of Image Identify in the Wolfram Language could be used for a wide variety of applications. A photo organizer, for example, could use the tool to sort all of your photos by their contents, rather than boring old things like dates and locations.
For now, though, go have some fun seeing exactly what Wolfram's Image Identification Project can and can't recognize over on their new website. And if you're interested in how the tool was developed, Wolfram's blog post on it is well worth a read.