Babar Khan Javed
Jul 3, 2018

Microsoft face-recognition tech gets some bias training

Microsoft has updated its facial-recognition technology to recognise various skin tones and associate more genders with search terms.

Users of all skin tones will now recognized by the system.
Users of all skin tones will now recognized by the system.

Face API via Azure Cognitive Services, the flagship facial-recognition technology at Microsoft, has been retrained to recognize a range of skin tones and genders. According to the announcement post, the upgrade reduces error rates for men and women with darker skin by up to 20%, with error rates for women reducing nine times. 

Working with experts on bias and fairness, Microsoft managed to model a gender classifier to get better results across all skin tones and recognise genders associated with search terms. The change mirrors similar steps taken by IBM in removing bias from facial recognition technology

"For example, if you do a web search for the word CEO, chances are you’ll get information about the senior leadership position in companies and organizations around the world, including a handful of images—most likely of men," said John Roach, editor of the AI blog at Microsoft. The results have more to do with SEO best practices that add necessary tags on images pertaining to the name, title, and company an image is associated with. It also has to do with the fact that under 5% of Fortune 500 CEO's are women.

Hanna Wallach, a senior researcher in Microsoft’s New York research lab has been looking into how best to surface search results which reflect gender-neutrality. Her team is rectifying the matter by incorporating diverse datasets in the development process in order to train systems in identifying a range of skin tones, gender and age groups. In doing so, Wallach is improving the classifier to produce higher precision results.

Related Articles

Just Published

1 hour ago

LinkedIn logs out: Is it the end of the road for ...

SOUNDING BOARD: Industry experts weigh in on the future for LinkedIn and other non-native tech platforms in China, as the business-networking site confirms it will shrink to a jobs board in the market.

1 hour ago

Japan continues to witness high ad fraud while ...

Grappling with high CPMs, Japan and Singapore report some of the highest fraud rates for IAS, while local brands pull rates down in Indonesia.

1 hour ago

Digital Media Awards 2022: Call for entries

The DMAs celebrate outstanding ideas, work, agencies, brands and talent in the Greater China region.

2 hours ago

BrandZ China: Tencent overtakes Alibaba; Kuaishou, ...

The top 100 most valuable brands in China gained 57% in brand value year-over-year, reaching $1.56 trillion, according to Kantar's latest ranking.