As photo retouching software becomes more accessible and prevalent, the models and celebrities who grace magazine covers sport slimmer bodies and better complexions than ever before. Motivated by the relationship between unrealistic, altered images and negative body image among readers, computer science professor Hany Farid and Eric Kee GR'13 have developed a system for quantifying the amount of retouching images receive.
The work published in the Proceedings of the National Academy of Sciences in November responds to studies that link retouching to issues like eating disorders, according to Kee.
"I think that body image and dissatisfaction is a significant problem in our society," he said. "The images that we see in newspapers and magazines kind of document what we think the standard of beauty is, and I think a rating system could kind of calibrate it back to reality again."
Farid and Kee's work also follows a recent push in Europe for legislation that would require advertisers and publishers to disclose image retouching. Such policies would simplify the issue, failing to distinguish photos cropped and lightened from those featuring more extreme "digital plastic surgery," Farid said.
To develop its model, the team assembled sets of original and retouched images and asked participants to rate them based on how seriously they thought the subjects' appearances had been altered. With this data, Farid and Kee generated a formula that quantifies the extent of digital retouching on a scale of one to five.
A group at the University of Amsterdam is currently evaluating the impact that this kind of rating system could have on readers' perception of printed photos and body image, according to Farid.
Recently, Farid and The New York Times performed a photo analysis on officially released photos of deceased North Korean dictator Kim Jong-Il's funeral procession and determined that the images had been altered, according to The New York Times.
While the official image showed an organized crowd, the original image revealed a more scattered arrangement of onlookers.
Farid analyzed the locations of parade participants in relation to their footprints to discern the accuracy of the photo, The Times reported.
Readers, especially young adults, often assume images in mass media have been altered, according to students interviewed by The Dartmouth.
"Whenever I see those shows about models and stuff, like America's Next Top Model,' you see those pictures of the girls before and after and they're totally different," Laura Bergsten '15 said. "I just kind of expect it."
Although Dartmouth Engineer magazine does not have an official policy on photo retouching, editor Karen Endicott of the Thayer School of Engineering said she only permits minor alterations, such as brightening, cropping and blemish removal, processes that Farid and Kee's participants supported.
"If I have a close-up of a student and the face is really blemished, I might even out a little bit of that, but we don't alter the meaning of a photo," Endicott said. "I think it's wrong to alter the meaning of a photo."
Although it could take years before his research code is developed into a software packet for commercial use, Farid and Kee's work has already had visible effects, Kee said.
"It's interesting to note that even without creating the software it's creating change politically and socially," Kee said. "Just by saying, Hey, we can do this,' we've raised awareness of the issue. That alone could drive change as well."
