Slashdot Mirror
Amazon Is Pushing Facial Recognition Tech That a Study Says Could Be Biased (nytimes.com)
An anonymous reader quotes a report from The New York Times: Over the last two years, Amazon has aggressively marketed its facial recognition technology to police departments and federal agencies as a service to help law enforcement identify suspects more quickly. Now a new study from researchers at the M.I.T. Media Lab has found that Amazon's system, Rekognition, had much more difficulty in telling the gender of female faces and of darker-skinned faces in photos than similar services from IBM and Microsoft. The results raise questions about potential bias that could hamper Amazon's drive to popularize the technology.
In the study, published Thursday, Rekognition made no errors in recognizing the gender of lighter-skinned men. But it misclassified women as men 19 percent of the time, the researchers said, and mistook darker-skinned women for men 31 percent of the time. Microsoft's technology mistook darker-skinned women for men just 1.5 percent of the time. For the latest study, [co-author of the study, Ms. Buolamwini, said] she sent a letter with some preliminary results to Amazon seven months ago. But she said that she hadn't heard back from Amazon, and that when she and a co-author retested the company's product a couple of months later, it had not improved. "It's not possible to draw a conclusion on the accuracy of facial recognition for any use case -- including law enforcement -- based on results obtained using facial analysis," Matt Wood, general manager of AI at Amazon Web Services, said. He added that the researchers had not tested the latest version of Rekognition, which was updated in November.
"Amazon said that in recent internal tests using an updated version of its service, the company found no difference in accuracy in classifying gender across all ethnicities," the NYT reports. The new study is scheduled to be presented Monday at an artificial intelligence and ethics conference in Honolulu.
In the study, published Thursday, Rekognition made no errors in recognizing the gender of lighter-skinned men. But it misclassified women as men 19 percent of the time, the researchers said, and mistook darker-skinned women for men 31 percent of the time. Microsoft's technology mistook darker-skinned women for men just 1.5 percent of the time. For the latest study, [co-author of the study, Ms. Buolamwini, said] she sent a letter with some preliminary results to Amazon seven months ago. But she said that she hadn't heard back from Amazon, and that when she and a co-author retested the company's product a couple of months later, it had not improved. "It's not possible to draw a conclusion on the accuracy of facial recognition for any use case -- including law enforcement -- based on results obtained using facial analysis," Matt Wood, general manager of AI at Amazon Web Services, said. He added that the researchers had not tested the latest version of Rekognition, which was updated in November.
"Amazon said that in recent internal tests using an updated version of its service, the company found no difference in accuracy in classifying gender across all ethnicities," the NYT reports. The new study is scheduled to be presented Monday at an artificial intelligence and ethics conference in Honolulu.
Software can be inaccurate. It can't be biased.
I'm looking out my window and it is sunny and low 80's. You should move somewhere better.
Doh!
>to be presented Monday at an artificial intelligence and ethics conference in Honolulu.
When you really want in depth information on AI and ethics, you HAVE to go spend a week in beautiful Hawaii. That's where all the.... uh.... stuff is.
that all black people look the same?
So the message is that the software is much more likely to be successful at apprehending guilty white people? Sorry for using a racist tag ("white") in my comment.
It does sound like it's strongly biased against white people and should be scrutinized carefully.
...when the concern about this tech is not that it exists but that it might be "unfair" to some artificial identity politics minority.
Lol idiots (journalists)
How well does it discriminate between the individuals? Law enforcement should focus on known criminals and the behaviour of the individuals rather than their gender, which should be more important for researchers and marketers.
Along with all the other morons who believe that local weather = global climate.
Where are we going and why are we in a handbasket?
the company found no difference in accuracy in classifying gender across all ethnicities
Maybe the spokesperson is clueless, but ethnicity is not race. Look at people in Cuba: some appear Black, some European, some Native American, many are mixed. But all are Hispanic ethnicity.
work by recording the light reflected from objects.
Darker-toned faces reflect less visible-wavelength light.
That's just physics, not racism.
So the amount of light, and ability to resolve contrasts, edges etc, would be less.
So the image classification task might be subject to more error.
Perhaps a different spectral range would work better?
Where are we going and why are we in a handbasket?
Why the hell would you WANT some fucked up bezos computer to identify you as you walk down the street?
I would say it is biased against bald beardless white men. Black women get misidentified at such a high rate the tech is worthless to identify them. That is a GOOD THING FOR BLACK WOMEN!!!
Iâ(TM)m not bald and now I am definitely keeping my beard. For once, a benefit to being a black woman: you dont get tracked and stored by bezos and his evil minions.
And any amazon engineers reading this who worked on the tech, SHAME ON YOU! You could have earned just as much working on tech that is not evil and not putting it straight into the hands of evil people and everyone with a dollar.
When a person recognises another person's face we usually mean to say that they've seen the person before and/or can possibly identify the person.
This slashdot article suggests that something else is meant here: gender and race recognition. Is that indeed the case? Are we asking law-enforcement systems to identify gender and race?
If so, to what end? To find people based that match often vague descriptions?
I'm probably being a moron for not realising this until now. All this time I thought they were just looking to match people on the street with photos of people that are "wanted".
I bet facial expressions make a big difference looking at still pictures
It does not matter what color, gender or religion you are. We can All be Bros. The important thing to do is divide the world up along the lines of thin crust pizza or thick crust pizza. All people who eat thick crust are destroying the planet and raping babies. They need to die.
Go thin crusy
Why is this a good think for Black women? It seems like they'd be falsely arrested at a higher rate than other people, assuming the tech is seeking matches for wanted criminals or terrorists.
Facebook would *never* promote a technology without thoroughly thinking through the implications. They are the pinnacle of corporate social responsibilty...
Come to think of it, that last part may actually be true.
Post may contain irony: discontinue use if experiencing mood swings, nausea or elevated blood pressure.
Obviously this technology needs to be banned from use until it misidentifies men as women as often as it misidentifies women as men. We can't allow anything that yields unequal results from ever being used.
I'm human and in 2019 I'm not always sure - is the AI being trained to recognize chromosomes or affect? This is not a troll, I am serious.
therefore this is a non-story.
Y'all look alike to the rest of us.
"Facial Recognition Tech That a Study Says Could Be Biased"
Is Facial Recognition Tech, supposed to be absolutely error free, or, how much error it is allowed to have exactly?
Is Facial Recognition Tech, always do less errors than any human doing the same job, or more?
If Facial Recognition Tech does errors, what is the harm of those errors exactly?
If a facial recognition tech identifies somebody as a criminal, are the police just go & arrest that person & put him/her to jail?
Nobody ever checks/verifies (first) whatever Facial Recognition Tech says?
They probably trained this on their full-time development staff.
They should have included warehouse staff, and then a double measure of cleaning/maintenance staff.
There. Fixed that.
Some people are under the mistaken impression that something wildly inaccurate and highly biased towards false positives wouldn't be sufficient to void Constitutional rights. The Supreme Court put that notion to rest when they ruled a cop merely needs to claim a dog trained to please him gave permission, and the 4th Amendment is gone. People who think inaccuracy is a good thing will be sorely disappointed. Well, they won't until the day comes when they're on the receiving end of police bullshit, but then they will.
Remember when Google's image algorithm classified multiple black people like gorillas and monkeys? YouTube's video suggestion algorithm did the same thing: After a report on a crime committed by black people, viewers were recommended a video of a baby gorilla born at a zoo.
This pattern recognition was completely unbiased. Without human political correctness, computers do think black people look like gorillas, and that's a harsh pill for many to swallow.
Am I surprised that black people are again on the raw end of computer logic? No. Computers think that black women look like black men a significant percentage of the time. Programmers now need to try and figure out how not to misgender black women (a really hard problem to solve in technology). I suggest that young non-black children might even have a similar struggle and error rate in black gender identification by facial photo.