Google's Latest Attempt to Reduce AI Bias: a Skin Tone Palette with 10 Different Shades

Tech giant Google announced this week that it plans to use a skin tone palette featuring 10 different shades of skin color for its apps to reduce bias in its AI system.

The Hill reports that tech giant Google announced this week that it will be implementing a new Monk Skin Tone (MST) Scale into its various products over the next few months, stating that the move was an effort to support inclusive products and research in the industry.

The MST scale was created by Harvard professor and sociologist Ellis Monk and was designed to be used in the development and evaluation of technology while representing a broader range of skin tones; something which AI machine learning models have struggled with in the past.

Pregnant man emoji in different skin colors (Emoji 14.0)

In a statement, Monk said:

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.

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 stated that the skin tone palette could give the tech giant a better understanding of how representation is viewed in the industry. The company stated:

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. For example, we use the scale to evaluate and improve the models that detect faces in images.

The MST is set to replace the Fitzpatrick Skin Type, a widely used scale that displayed six skin color types but drew criticism for allegedly underrepresenting people with darker skin tones.

COMMENTS