Flickr adds new feature to search images based on visual similarity

Flickr now lets you search image that shares the same shape, composition, colour, or content category. Yes, Yahoo is updating Flickr with a new feature that allows users to search for images that are visually similar to a single given image.

In a blog post, Flickr wrote, "today we are introducing similarity search on Flickr. If you hover over a photo on a search result page, you will reveal a “…” button that exposes a menu that gives you the option to search for photos similar to the photo you are currently viewing."

Flickr uses deep neural networks to analyse the photos as they’re uploaded to the server. Instead of only showing images containing the same object, Flickr’s visually similar image search — internally called the “similarity pivot” — fetches photos that have certain definite connections, like shape, colour, or general categories.

“The similarity pivot is a significant addition to the Flickr experience because it offers our community an entirely new way to explore and discover the billions of incredible photos and millions of amazing photographers on Flickr. It allows people to look for images of a particular style, it gives people a view into universal behaviors, and even when it ‘messes up,’ it can force people to look at the unexpected commonalities and oddities of our visual world with a fresh perspective,” Yahoo software engineer Clayton Mellina wrote in the post.

The system is not perfect, however. Flickr says users may come across photos that may not share anything but similar shapes. The company says “even when ‘messes up,’ it can force people to look at the unexpected commonalities and oddities of our visual world with a fresh perspective.”

Other companies are also trying to integrate deep learning into their services. Facebook is seeking to generate image captions automatically for screen readers to help blind users to understand the post. Microsoft is also working on similar technology on Word and PowerPoint.