Technology giant Google announced this week that it plans to use a skin tone palette with 10 different skin colors in its apps to reduce bias in its AI system.
Hill reported this week that tech giant Google will launch the new Monk Skin Tone (MST) scale across its various products in the next few months, and the move is an attempt to support inclusive products and industry research. .
The MST scale was created by Harvard professor and sociologist Ellis Monk and is intended for use in the development and evaluation of technology that represents a broader spectrum of skin colours; It’s something AI machine learning models have struggled with in the past.
Expression of a pregnant man with different skin colors (Emoji 14.0)
Monk said in a statement:
In our research, we found that people often feel they are racially classified, but there is all this heterogeneity in ethnic and racial categories.
And many categorization methods, including the skin color scales of the past, do not take this variation into account. This is where a lack of representation can come in… We need to adjust the way we measure things so that the people they represent can feel it.
Google said the skin tone palette could help the tech giant better understand how representation is viewed in the industry. The company said:
The size of the MST will help us, and the tech industry as a whole, build more dataset representations for us to adapt and evaluate AI models for fairness, resulting in features and products that work best. For example, we use scales to evaluate and improve models that see faces in photographs.
The MST was designed to replace the Fitzpatrick Skin Type scale, a widely used scale that displays six skin color types, but has come under criticism for allegedly under-displaying darker skin tones.
Read more about Hill here.
Source: Breitbart
