app

'The Dress' Fascinates Vision Scientists

— Clues to neurology, psychology of visual perception from 2015 internet sensation

MedpageToday

ANAHEIM, Calif. -- Remember "The Dress," the Internet sensation that divided the world last year? Some of us see a blue-and-black dress, while others swear it's gold-and-white.

And then there's . He looks at the dress and sees blue ... and burnt gold. "I'm an outlier," Rabin said, and one with plenty of company: About 8% of people don't see one of the two predominant color combinations.

What's going on? How can so many people see so many variations? In a new study, Rabin and colleagues report they've discovered clues that may help explain why we see The Dress so differently. They've found signs that macular pigments and blue cones may play roles in addition to neural processing.

"We haven't solved it. But we're emphasizing the need to consider not just higher-level processing but also front-end factors that may predispose us to see it one way or another way," said Rabin, of the University of the Incarnate Word's Rosenberg School of Optometry.

Rabin discussed his team's findings during a presentation at the here, and in an interview with app. The research was also .

The different perceptions of the colors of "The Dress" have baffled vision scientists. "We hadn't seen this dichotomy in color perception before," Rabin said.

One popular theory is that our brains make assumptions about the colors of the dress based on our perceptions of how the dress is illuminated in the photo. But Rabin said that explanation isn't enough.

For the study, Rabin and colleagues recruited 39 visually normal observers (mean age 32, SD 10; 20 men) and showed them a photo of the dress.

Three subjects perceived blue and gold like Rabin does. They were not included in the analysis. Of the others, 19 saw blue/black and 17 saw gold/white. (The audience members at Rabin's conference presentation were similarly divided, about half and half, between the two options.)

There was no difference in perception even when the researchers took away the context of the dress photo by cropping it.

The researchers found evidence that those who perceived gold/white had denser macular pigment. "Greater pre-retinal absorption of blue light by denser macular pigment may predispose observers to perceive white/gold," Rabin said.

The researchers also performed an analysis of the image "and found there's a very strong blue cone contribution," Rabin said. "My feeling is that in most real-world images that we encounter, we probably don't get this disproportionate stimulation of blue cones. Maybe our visual systems don't know how to interpret it when a wrench gets thrown into the works."

The researchers also found evidence that people who perceive the dress as gold/white may need to go through an extra step in neural processing compared to those who see blue/black. "Processing is probably taking place in different parts of the visual system" in these two groups of people, he said.

Rabin isn't satisfied that these factors fully explain a phenomenon that's so intrigued visual scientists that an entire issue of the Journal of Vision will be devoted to "The Dress."

"I don't have a perfect answer," he said.

But Rabin said one thing is clear: We need to look beyond back-end neural processing when visual perceptions go out of whack. "We hope that this can be applied to atypical behaviors and perceptions in senescent brain disease, Alzheimer's and mild TBIs [traumatic brain injuries] with unusual symptoms," he said. "Go forward from the brain."

  • author['full_name']

    Randy Dotinga is a freelance medical and science journalist based in San Diego.

Disclosures

Rabin reported no relevant disclosures. No specific funding was reported.