Google to use 10 shades in skin tone palette to reduce bias in AI
Google on Wednesday announced that it will use 10 shades in skin tones palettes for its gadgets and apps in an effort to reduce bias in its AI system.
In a statement, Alphabet Inc. said the new Monk Skin Tone (MST) Scale will be incorporated into various Google products over the coming months, saying the new scale will support inclusive products and research across the industry.
The MST scale, created by Havard professor and sociologist Ellis Monk, was designed to be easy to use for the development and evaluation of technology while representing a broader range of skin tones, adding that it’s more representative than other skin tone scales in the industry currently.
“In our research, we found that a lot of the time people feel they’re lumped into racial categories, but there’s all this heterogeneity with ethnic and racial categories,” Monk said in a statement.
“And many methods of categorization, including past skin tone scales, don’t pay attention to this diversity. That’s where a lack of representation can happen … We need to fine-tune the way we measure things, so people feel represented.”
Google also said that the skin tone palette can help the company better understand how representation is viewed in the industry and evaluate how a product or feature works across a range of skin tones, the statement said.
“The MST Scale will help us and the tech industry at large build more representative datasets so we can train and evaluate AI models for fairness, resulting in features and products that work better for everyone — of all skin tones,” the company said in its statement. “For example, we use the scale to evaluate and improve the models that detect faces in images.”
The move comes as MST is set to replace the Fitzpatrick Skin Type, a widely popular scale known for its six skin color types, Reuters reported. The Fitzpatrick scale saw some criticism, with analysts saying the scale underrepresented people with a darker skin tone.