EyeEm’s Software Knows What’s Pretty
A person doesn’t “decide” a photo is beautiful—the reaction is emotional, instinctive. So when Appu Shaji began trying to duplicate the process with code four years ago, building a program to rank photos by their beauty, he failed. A computer imaging researcher at EPFL, an engineering college in Lausanne, Switzerland, Shaji designed his software to judge photos based on composition, color saturation, and perspective, but couldn’t teach it to identify the best-looking ones. “With hard rules, I wasn’t able to solve the problem,” he says.
So Shaji rounded up Berlin’s best photographers and artists and asked them to judge the same images. By tweaking his code based on their input, he says, he slowly made the software good enough to match the taste of a person instead of a machine. Last year, Berlin software startup EyeEm bought Shaji’s company, Sight.io, for an undisclosed amount and made him the head of research and development.
Shaji’s technology is at the heart of EyeEm’s online photo market, a cross between Instagram and the reusable stock photos of Getty Images. The company, which on April 16 announced an $18 million round of funding from investors including PayPal co-founder Peter Thiel, may be the closest to solving one of the software industry’s biggest technical problems: how to quickly and effectively sift through the trillions of images flooding the Internet. Companies from Airbnb to Yahoo! are investing in efforts to develop better image search features. So far, most software simply identifies the subject of an image—“cat,” or maybe “cat sleeping.” Shaji’s image-ranking software goes a step further.
The 13 million registered users of EyeEm’s app upload and edit photos, add filters, and share them with followers. Ad and media outlets can use EyeEm’s search engine to scour about 50 million photos for the prettiest “East Coast sunset” or “beach with no people” and license a photographer’s work through the app. Commercial use online, which can include a print run of up to 250,000 copies, costs $20 a photo; unlimited print or TV use is $250. EyeEm and the photographer split revenue 50-50, says spokesman Drew Olanoff. He wouldn’t disclose total revenue.
“EyeEm’s technology could expand the commercial market for image sales by an order of magnitude,” says Andrew McCormack, a partner at investor Valar Ventures who joined EyeEm’s board after its latest funding. “Everyone’s a photographer. The number of images and the quality of images is going up logarithmically. How can you make everyone a seller and a buyer?”
In some cases, companies work with EyeEm to rally photographers for contests. The process begins with a photo request from a brand, such as for shots of people in different countries eating a certain kind of ice cream. EyeEm’s beauty filter sifts user submissions to find the ones it deems the best.
Among advertisers, “there’s this huge demand for authenticity,” says Jason Whitmire, a partner at investor Earlybird Venture Capital. “If you go at the regular marketplaces like Getty or Shutterstock and you search for a picnic, you’re going to get all these staged, almost fake photos.” Shutterstock Chief Product Officer Catherine Ulrich says her company’s clients often prefer photos that are staged and less arty. Getty has linked up with EyeEm to offer subscribers access to about 300,000 of EyeEm’s photos.
Yahoo researcher Alejandro Jaimes says he’s trying to build software that can help people pick their most flattering profile pictures. Airbnb engineer Hector Yee says his photo research shows that people looking for rooms to rent are most interested in bedroom shots. For now, though, EyeEm has a head start on digitizing good taste. The four-year-old, 55-employee company is working to refine its systems as its library grows, says Shaji. “We have just scratched the surface,” he says.
The bottom line: Online stock photo marketplace EyeEm has raised $18 million for a cross between Getty Images and Instagram.