Shutterstock recently announced the launch of it’s Reverse Search for Video, a search and discovery feature powered by computer vision technology. Users can simply upload an image of their choosing and the AI-powered algorithm will instantly provide videos similar in look and feel to the original, overcoming the limitations of keyword search.

This could be a wonderfully helpful tool for videographers and video editors, as they claim it’ll make missing shots a thing of the past. It is now much easier to find footage resembling — or shot in — a certain location as well as scenes that feature a certain light temperature, wall color, or object.

What is Reverse Image Search?

In the simplest terms, Shutterstock is using “machine learning” AI to break down it’s library of over 280 million images and 15 million video clips at a pixel level, identifying key features. Then once it’s recognized what’s inside each image, it visually searches the database (from your submitted image) to find others that match its basic composition, returning you results that should match the image/video you’re looking for.

It’s been working pretty well for Shutterstock for the last few years in terms of photo searches, and now with its adaptation for video, it effectively searches the clips frame by frame to find results that match the image or screenshot you submit

For instance, as a test I submitted an image I captured while in Death Valley of the Eureka Dunes

Death Valley - David J. Crewe
Eureka Dunes – Death Valley

What returned was a variety of sweeping landscapes, dunes, and mountains as seen here;

Shutterstock Reverse Video Search Results - Death Valley - SLR Lounge
Shutterstock Reverse Video Search Results – Death Valley

[Related Reading: Stock 4K? Shutterstock Quietly Launches 4K Support]

Given how generic the image I uploaded as a test was, i’m actually pretty impressed with the resulting video samples it returned! You can effectively throw in anything into the search and while it may take a while to analyze, it’ll return with some pretty accurate results for your video clip needs. Regardless of your stance on stock photography/videography, this is still a pretty impressive feature for people putting together a video project. Definitely a MASSIVE time saver now that you can simply drag and drop a screenshot of your existing footage to find clips that are similar in nature and available for licensing.

Personally, if I were a creative director or producer looking to finish up a film or commercial project, this would be incredibly useful, and a massive time and cost saving tool. What do you think of this? Will you use this for your video projects? Let us know in the comments below.