Image and music content provider Shutterstock has launched a new visual search tool that doesn’t need keywords to find relevant images for clients.
The new search and discovery features powered by its own custom-built convolutional neural network.
Reverse image and advanced visually similar search capabilities for images fully launch today and Shutterstock will soon launch visually similar discovery for video.
The first tool using the new system, dubbed ‘Reverse Image Search’ provides an innovative alternative for customers as compared to using keywords to search for images in Shutterstock’s collection.
Users can simply upload a photo of their choosing from Shutterstock’s collection or from another source, and the algorithm will detect and then, in an instant, provide images similar in look and feel to the original, overcoming the limitations of keyword search.
“We’re continuing to build industry leading technology to improve our offering. With a collection as vast as Shutterstock’s, the importance of being able to surface exactly what a customer needs with advanced search and discovery tools is essential to our continued success.” said Shutterstock’s founder and CEO Jon Oringer. “Doing this in video is a breakthrough and as the technology continues to learn and recognize what’s inside an image or a clip, it promises more possibilities. We know we’ve only scratched the surface in how we use this deep machine learning to better understand and serve our customers.”
Computer vision is the ability for a computer to break an image down into its primary characteristics, both visually and conceptually that can be represented numerically.
The technology relies on pixel data within images – rather than metadata collected through keywords and tagging – to help identify and surface relevant content.
Shutterstock’s computer vision team, that built and tested this technology entirely in-house, was put together over a year ago to solve the challenges associated with visual search.