Most engineers are white — and so are the faces they use to train software – Recode

2017/01/19 Leave a comment

Not terribly surprising but alarming given how much facial recognition is used these days.

While the focus of this article is with respect to Black faces (as it is with the Implicit Association Test), the same issue likely applies to other minority groups.

Welcome any comments from those with experience on how the various face recognition programs in commercial software such as Flicker, Google, Photos etc:

Facial recognition technology is known to struggle to recognize black faces. The underlying reason for this shortcoming runs deeper than you might expect, according to researchers at MIT.

Speaking during a panel discussion on artificial intelligence at the World Economic Forum Annual Meeting this week, MIT Media Lab director Joichi Ito said it likely stems from the fact that most engineers are white.

“The way you get into computers is because your friends are into computers, which is generally white men. So, when you look at the demographic across Silicon Valley you see a lot of white men,” Ito said.

Ito relayed an anecdote about how a graduate researcher in his lab had found that commonly used libraries for facial recognition have trouble reading dark faces.

“These libraries are used in many of the products that you have, and if you’re an African-American person you get in front of it, it won’t recognize your face,” he said.

Libraries are collections of pre-written code developers can share and reuse to save time instead of writing everything from scratch.

Joy Buolamwini, the graduate researcher on the project, told Recode in an email that software she used did not consistently detect her face, and that more analysis is needed to make broader claims about facial recognition technology.

“Given the wide range of skin-tone and facial features that can be considered African-American, more precise terminology and analysis is needed to determine the performance of existing facial detection systems,” she said.

“One of the risks that we have of the lack of diversity in engineers is that it’s not intuitive which questions you should be asking,” Ito said. “And even if you have a design guidelines, some of this stuff is kind of feel decision.”

“Calls for tech inclusion often miss the bias that is embedded in written code,” Buolamwini wrote in a May post on Medium.

Reused code, while convenient, is limited by the training data it uses to learn, she said. In the case of code for facial recognition, the code is limited by the faces included in the training data.

“A lack of diversity in the training set leads to an inability to easily characterize faces that do not fit the normal face derived from the training set,” wrote Buolamwini.

She wrote that to cope with limitations in one project involving facial recognition technology, she had to wear a white mask so that her face could “be detected in a variety of lighting conditions,” she said.

“While this is a temporary solution, we can do better than asking people to change themselves to fit our code. Our task is to create code that can work for people of all types.”

Source: Most engineers are white — and so are the faces they use to train software – Recode

Filed under Multiculturalism Tagged with ,

The End of Second Languages Is Near – The Daily Beast

2016/05/30 Leave a comment

The strengths and limitations of electronic translation tools:

Remember all those warnings that learning a second language would be a great asset in your future? It seems that years after your regrets about not paying attention and learning Spanish, technology has finally caught up and proven them wrong.

Tech company Waverly Labs is taking pre-orders for the kind of device you’d have seen in a space epic 30 years ago: a tiny smart earpiece, called “Pilot,” that real-time translates one language into another. Speak Spanish, hear English. Get it?

For decades, it’s been the work of science fiction to imagine a tiny earpiece translating someone’s words from another language to your own in real time. But in the last few years we’ve breached plenty of technological barriers to quick translation. Think of the dozens of languages available with Google’s translation tool, or the multiple apps that can translate signs, menus, and text in real time by using the camera on a smart device.

And all this without a live translator. This is just the latest in a series of technological advances in translation. The result: knowing the other language is unnecessary if you have the right tools.

For years it’s been a major boon in business to know a second language—and for the sake of relationships, it may still be. But it looks like in a few years you’ll be able to attend a German cocktail hour and know what’s being said, or make that trip to France and understand directions.

But the idea that a small earpiece and a couple of phone apps can turn that intimidating interaction into something casual (and that you can one day talk to Italian, Chinese, Indian, and Russian businesspeople without learning a word of their language) makes this hurdle a lot shorter.

That’s good news, especially for the U.S., because Americans aren’t great with second languages. About a quarter of Americans speak a second language conversationally, while the rest of the world’s polyglot numbers go above 50 percent.

It’s not a true closure to the gap though. There’s still interpersonal significance to the idea that you can speak to someone in their native language without some apparatus doing the heavy lifting—and if that person doesn’t speak your language, you can’t get by on one earpiece alone.

Furthermore, those same international businesspeople that you might envision working for or with want you to be fluent, not to buy an earpiece. There are countless studies and data points out there showing how much more valuable you are to a company as someone who can interact in two or more languages, and that’s not going to change within five or 10 years of a new Bluetooth earpiece hitting the market.

Still, it’s a lot less work than something like Rosetta Stone, because it doesn’t actually require you to learn anything—just to plug and play. And while multimillion dollar deals require fluency, those little emergency interactions on the street when you’re totally lost in a foreign city, or have your passport stolen, or don’t understand the taxi driver—those are all great occasions where technology can take over.

With the ubiquity of cellphones, particularly smartphones, increasing worldwide, it doesn’t look so touristy to hold up your phone to a sign to translate it. And tossing an earpiece in before jumping on a conference call might make your life a lot easier if there are non-English speakers speaking.

The final question then is one of perception: How bad does it look to let technology do the work? You would never write a formal business proposal and let Google do the translating work: it would lose grammar, craft, and it would miss idioms. Maybe then a few months or years of language study is good just so you don’t have one of those embarrassing misunderstandings where you tell someone they’re pregnant instead of smart.

Waverly Labs shows the Pilot smart earpiece as bridging the gap between an affectionate couple that experienced love at first sight, but never shared a common language. And as wonderful as the “thank god for technology” moment is, wouldn’t you rather be the kind of romantic that learns another person’s native tongue for them?

 Source: The End of Second Languages Is Near – The Daily Beast

Filed under Multiculturalism Tagged with , ,