Google Image Swirl organizes image search results based on their visual and semantic similarities and presents them in an intuitive exploratory interface. How does this work? Google Image Swirl combines a variety of image similarity features with additional metadata about the images to build a hierarchy of clusters of image search results. The clustering algorithm is fully automatic, and far from perfect. In most cases, computer vision cannot tell us what actually present in an image, and without this semantic information, mistakes will happen. We welcome your feedback about cases which work well and which need more work. Image Swirl currently works for more than 200,000 queries and we plan to include more queries in the future. Available queries will auto-complete as you start to type in the search box, similar to Google Suggest.
 
Bron: image-swirl.googlelabs.com