Tuesday, November 29, 2011

Photo Retouching Detector of the Day

via TDW Geeks by Cheezburger Network on 11/29/11

Photo Retouching Detector of the Day
Photo Retouching Detector of the Day: When Dartmouth professor Hany Farid says "this looks 'shopped," he's got the science to back it up.
Farid and PhD student Eric McKee have developed an algorithm that measures the level of retouching in magazine photos, and displays the results as a heat map.
The researchers developed the algorithm by looking at anatomically impossible images pulled from Photoshop blunder sites. They fine-tuned it using crowdsourcing, giving people 70 sets of before-and-after photos and asking them to rate the difference on a scale of 1 to 5.
Farid thinks magazines should show his heat maps alongside their grotesquely 'shopped images, because revealing how unrealistic the photos are would encourage them to reduce retouching.
Or, in Britain, it could help courts decide which ads to ban.
[economist]

No comments:

Post a Comment