Google has launched a new Labs experiment dealing with image search. The feature is called Image Swirl, and attempts to deliver the user image search results in a different and more visual interface. Google describes Image Swirl:
Google Image Swirl organizes image search results into groups and sub-groups, based on their visual and semantic similarity and presents them in an intuitive exploratory interface. Think about Image Swirl as kind of a visual image version of the wonder wheel. The wonder wheel performs a very similar task with text. This works with actual images (not just image queries). Here’s what it looks like:
Google says to try Image Swirl to resolve an ambiguous query visually, such as “apple,” “jaguar,” or “beetle.” They also suggest using it to explore different visual perspectives on queries like “Eiffel Tower,” “beach,” or “impressionism.” You can only use the feature for select queries at this point. It’s not just like performing a regular image search. They give you a list of possibilities in addition those mentioned above, but it is not limited to those. I tried searching for several band names, for example, and they all seemed to work.
After that quick test run, I could actually see this being useful in a shopping scenario. For example, when I search for the band “Converge,” I get one result that shows a t-shirt, and when I click on that, it brings up other t-shirt results from various sites. If I was looking to buy a new Converge t-shirt, that could come in quite handy.
Google Labs experiments are of course just that – experiments. Some of them blossom into full-fledged Google products. Others sit there and collect dust. I can already see a great deal of potential in this one, however, and based on Google’s use of the Wonder Wheel in its search options, I could easily see this one making a jump to the image search options.
What do you think of the Image Swirl? Would you use this if it was a feature of Google Image Search? Share your thoughts.
Related Articles:
> Google “Similar Images” Feature Goes From Labs to Actual Feature
