AI has the potential to revolutionise (perhaps even to terminate) the working lives of millions. But in the meantime, scientists and academics from Oxford University are using it to look for celebrity lookalikes in one of the world’s largest digital collections of art images.
The inventors of Pholio, an intelligent photo album with built-in advanced image search capabilities, have demonstrated the platform’s capabilities by training it to find celebrity lookalikes from The Witt Library (part of The Courtauld Institute), which holds a collection of western art from 1200 to the present day.
Among the celebrities whose likenesses they have found in old paintings are Taylor Swift, Cher, Donald Trump, Simon Cowell, Paul Hollywood, Richard Branson and Hugh Jackman.
Pholio, developed by Pimloc, is a small albumshaped box that sits on a living room shelf and uses artifcial intelligence to recognise patterns relating to more than 20,000 search terms. In this way, users can quickly and easily search their image collections for pictures of, say, eggs, cars or meerkats.
Founder Simon Randall said: “When we tested the search capabilities out on the Courtauld Collection, we were absolutely amazed at the speed and the accuracy of what we’d created. Most of us have tens of thousands of pictures these days, thanks to smartphones. Keeping track of what we have and hunting down the thing we are looking for is now an instantaneous exercise. Pholio means no more laborious scrolling through thousands of pics.”
Pholio can search for images on any connected device and can also access Google Images, Facebook and Instagram collections and download them for safe keeping. Tags are applied instantly to all images added to a collection.