Flickr today received a new update that will allow users to search for images based on visual similarity. Now, you’ll be able to pick an image from a search result and find more photos that share the same shape, composition, color, or content category.
In a blog post, Flickr says it uses deep neural networks to analyze the photos as they’re uploaded to the server. This helps to define the contents of the image, then flag whether or not it’s safe for work. It’s similar to what Facebook recently launched last month where users can search for photos based on their content, without requiring the uploader to manually add tags.
Flickr admits that its computer vision technology is not perfect, and users may come across photos that may not share anything but similar color tones. The company excuses the behavior by noting that “even when [the tool] ‘messes up,’ it can force people to look at the unexpected commonalities and oddities of our visual world with a fresh perspective.” Okay then.
Flickr may not be the first place you go to when trying to search for images, but it is a helpful tool for finding Creative Commons-licensed photos that fit your exact visual needs. The Similiarity Pivot tool is available now.