The company said it is new Monk skin tone scale replaces a defective six-color standard known as the Fitzpatrick Skin Type, which had become popular in the tech industry for evaluating whether smartwatch heart rate sensors, artificial intelligence systems, including facial recognition and other offerings, exhibit color distortion.
Technology researchers recognized that Fitzpatrick underrepresented people with darker skin. Reuters exclusively reported last year that Google was developing an alternative.
The company teamed up with Harvard University sociologist Ellis Monk, who studies colorism and had felt dehumanized by the cameras failing to detect his face and reflect the tone of his skin.
Monk said Fitzpatrick is great for classifying the differences between lighter skin. But most people are darker, so he wanted a ladder that “does a better job for most of the world,” he said.
Monk through Photoshop and other digital art tools curated 10 tones – a manageable number for people to help train and evaluate To the systems. He and Google surveyed around 3,000 people in the United States and found that a significant number said a 10-point scale matched their skin as did a 40-tone palette.
Discover the stories of your interest
Tulsee Doshi, head of product for Google’s AI team, called the Monk scale “a good balance of being representative and being manageable.”
Google is already applying it. Google Image beauty searches such as “bridal makeup” now filter results by Monk. Image searches such as “cute babies” now show photos with different skin tones.
The Monk scale is also implemented to ensure that a range of people are happy with the filter options in Google Photos and that the company’s face matching software is not biased.
However, Doshi said problems could seep into products if companies don’t have enough data on each of the tones, or if the people or tools used to classify the skin of others are affected by differences in lighting or personal perceptions.