Google has introduced more diverse skin tones to develop its artificial intelligence systems.
The new Monk Skin Tone Scale, named after Harvard professor Dr. Ellis Monk, has 10 skin tones.
Google says it will replace outdated skin tone grades that favor pale skin.
The tech company claims it will be used to improve products like Search and Photos.
What is a skin scale?
machine learning, a Artificial intelligence is used by many technologies, including cameras that recognize faces to unlock phones or automatically categorize photos.
But to get there, researchers need to train the technology to recognize a wide range of people.
To do this, they used something called skin scales. One of the most popular skin scales is the Fitzpatrick scale.
The Fitzpatrick scale was first put together in 1975. Its original use was to classify the response of different skin types to UV rays. Divide the skin into six shades.
Tech companies now use the Fitzpatrick scale to classify people in images and measure whether devices like facial recognition systems or smartwatch heart rate sensors work on all skin tones.
But many say the Fitzpatrick scale, which divides the six skin types into four different “white” skin categories, isn’t diverse enough.
“If you say, I tested my model to be fair to make sure it works on darker skin tones, but you’re using a scale that’s not representative of most people with those skin tones, and you don’t know how well it will work,” Google AI expert Xango Eyeé said.
What is the Monk Skin Tone Scale used for?
The Monk scale, announced in early May, has ten skin tones, which the company claims are representative of the wider population.
It’s already being used to filter results in Google searches — makeup image searches, for example, will show more different images. There are also some “true tone” photo filters, which Google says work better with dark-skinned images.
Google says it will eventually use the new scale across all of its products.
If other developers choose to adopt the Monk scale, it can also be used elsewhere.
Google has open-sourced the Monk Skin Tone Scale classification system – which means other companies can use it – in the hope that it will eventually replace the most popular skin tone scale in use today.
Why does this matter?
If technology isn’t developed using diverse data, it won’t work for everyone, regardless of their skin color.
But it’s not just a technical issue.
It has recently been recognized that many products do not represent different skin tones well.
In the past, many companies had a very limited or one selection of skin tone products, which were often designed for white skin tones.
crayon, plaster and Bodysuits are just some of the things that have become more inclusive in recent years.
Many find it frustrating that these changes took so long to happen.