Google Image Swirl is a very interesting approach Google is experimenting with to improve the searchability of images. Google Image Swirl organizes the Google Image Search results into their visually and semantically similar properties. As a product of Google Labs, Image Swirl is still in a demo phase and is not fully functional, but it still allows you to see the impressive twist on image search. From the Google Image Swirl Help page, there are over 200,000 queries currently supported that you can use with more coming on the way.
We can see a good example of how it works and what it does by doing a simple query for ‘umbrella’. Let’s say I need an icon or something like that. Your first results come back and it looks a lot like the results for the normal Google Image Search, except you might notice “stacks” of images underneath each result.
If you click on a stack, the Flash powered results page will “zoom” into the pile and fan out the results. The grouping is made by similarities like color and in this example of the umbrella, by orientation. You can keep zooming in to find a group that’s attributes match what you’re looking for.
From the same results page, you can see how the blue umbrellas were clustered under this stack instead of the red umbrellas in the stack that was clicked on in the previous screenshot. These umbrellas also tend to be pointing to the up and left.
When you get down to a single image and click on it, you’ll see that it takes you to a landing page where you can find the original image similar to how the current, mainstream Google Image Search works.
For a video demonstration, click the image below to play the umbrella example that I captured.
[youtubehq KNUzmOv2krc Video Demo of Google Image Swirl]
You can find out more information at the Google Image Swirl help page. Check out the Google Image Swirl demo product here: Google Image Swirl
Comment on this post in our forums
Related posts:
Comments are closed, but trackbacks and pingbacks are open.