Seeing a reflection of yourself in the world around you in real life, media, or online is essential to how you perceive yourself. We know that image-based technologies on the web have historically left people of colour feeling overlooked and misrepresented. Last year, we announced as our effort to improve the representation of diverse skin tones and diversity in image across. that’s why Google launces new ranking Signal : Diversity in Image
Google has reiterated our commitment to diversity in image and improving representation across our products. They have partnered with Harvard professor and sociologist Dr. Ellis Monk to release a new skin tone scale that includes the spectrum of skin tones reflected in society. For over ten years, Dr. Monk has studied how skin tone and colorism affect people’s lives. Dr. Monk’s research culminates in the Monk Skin Tone (MST) Scale, a 10-shade scale that will be incorporated into various Google products. Google aim to ensure that the scale supports inclusive products and research across the industry.
The scale is designed to develop and evaluate technology while representing a broader range of skin tones. Our research found that among participants in the U.S., people saw the Monk Skin Tone Scale as more representative of their skin tones than the tech industry standard. This resonates with people with darker skin tones.
In addition, we found that people feel lumped into racial categories, but there is heterogeneity between ethnic and racial categories. And many categorization methods do not pay attention to this diversity. We need to refine how we measure things to provide diversity in image so that people feel represented
improving approach to skin tone can help us better understand representation in imagery and evaluate how a product or feature works across various skin tones. This is vital for computer vision, an AI that enables computers to identify and understand images. When not intentionally developed and tested to include a broad range of skin tones, computer vision systems have not performed well for people with darker skin. The MST Scale will help develop more representative datasets so we can program and assess AI models for fairness, resulting in features and products suitable for all skin tones.
Improving skin tone representation in Search
Using the MST Scale, Google has introduced new features to make it easy for people of diverse backgrounds to find results relevant and helpful to their needs. For example, when you search for makeup-related queries in Google Images, there is an option to refine your results by skin tone. So when looking for “everyday eyeshadow” or “bridal makeup looks”, it is easy to find results relevant to your needs.
Online labelling of contents is crucial to how our systems provide relevant results. We are developing a standardized way to label web content to enable content creators, brands, and publishers to label their content with attributes like skin tone, hair color, and hair texture to easily understand search engines and other platforms.
The MST Scale will be used to enhance Google Photos. Last year, they have introduced an upgrade to the auto-enhance feature in partnership with professional image-makers. Google is launching a new set of Real Tone filters designed to work well across skin tones and evaluated using the MST Scale. We worked with Kennedi Carter and Joshua Kissi to evaluate, test and build these filters. These new Real Tone filters offer a variety of looks to reflect diversity in image. Real Tone filters will be released on Google Photos across Android, iOS and the web shortly.
We are releasing the Monk Skin Tone Scale to enable others to use it in their products and partner with us to learn together. We encourage your feedback to drive more interdisciplinary research and progress together. We will continue collaborating with Dr. Monk to evaluate the MST Scale across different regions and product applications and improve it to make sure the scale works for people worldwide. And we will continue with our efforts to ensure Google’s products work more effectively and better for everyone who engages with our products.
Article and Image Source : Google Article
Read the main Article from here : https://blog.google/products/search/monk-skin-tone-scale/