Computers Outperform Humans at Recognizing Faces

seven of five writes "According to the recent Face Recognition Grand Challenge, The match up of face-recognition algorithms showed that machine recognition of human individuals has improved tenfold since 2002 and a hundredfold since 1995. 'Among other advantages, 3-D facial recognition identifies individuals by exploiting distinctive features of a human face's surface--for instance, the curves of the eye sockets, nose, and chin, which are where tissue and bone are most apparent and which don't change over time. Furthermore, Phillips says, "changes in illumination have adversely affected face-recognition performance from still images. But the shape of a face isn't affected by changes in illumination." Hence, 3-D face recognition might even be used in near-dark conditions.'"

ORLY?

Who's this then :-)

Re:ORLY?

> Greetings, Professor Falken. Would you like to play a game?

in other news

computers outperform humans on math equations

Re:in other news

Right.

Recognition tasks are almost all inductive in nature, where performance on math is deductive. Human induction pretty well spanks machine induction at most of the things we take for granted - like recognizing and decoding faces, voices, speech, the sound of your walk, etc., etc., etc. The thing computers do least well is infer what bits of information are most important. We seem to excel at that.

Despite what the findings say, I stand by the faces thing. It sounds like the recognition algorithms got high-resolution 3D scans of human faces as input. Wake me when they can do as well as a human with low-resolution 2D scans.

That being said, it's great to see progress in this area. I can't wait until someone has to lop off my head and carry it with them in a plastic bag in order to break into my workplace. It's more grisly than taking a thumb, but much less likely to happen... I think...

Re:

I concur. Computers can outperform people under experimental conditions where the data provided to the computer is exactly what is required to perform the job at hand. Show me a computer that can recognize a person from a brief glimpse of the the back of

Re:in other news

The problem becomes how do you "flash" an image at a computer. A computer has a perfect memory. So you can't compare humans and computers in this way. A computer could completely memorize millions of faces, or even all the faces in the world, given enough storage space. 6 Billion people x 1 MB (exaggeration) per picture is only 6 petabytes. It's a lot of data, but not out of reach. So if computers get good enough at recognizing faces, it could become a useful too in security. Think about the security guard sitting at the front desk of 20 story building. Do you think he could identify every person who walks through those doors. Would he know if you were just using a stolen security badge?

Re:in other news

I can't wait until someone has to lop off my head and carry it with them in a plastic bag in order to break into my workplace.

I can wait - frankly if someone really wants to break into my office they can have my badge.

Great, now commercialize it..

It's really annoying how much of this research never gets turned into product.. or, worse yet, it gets embedded in some proprietary piece of shit hardware instead of being released as a reusable component. I'd love to add some good facial recognition to my pet robot, but I'm not buying your watt sucking camera.

Re:Great, now commercialize it..

It is being turned into product. Has you flown into the US from overseas recently? They have (and use) fingerprint scanners, cameras and facial recognition software running in US Customs. Sure right now only people with foreign passport have to scan in and back out when they go through customs, but the cameras are right in your face in the US passport lines as well. This may be very new, I first saw them this last Monday in JFK. But apparently this has been going on for awhile.

Re:

Gone through them about 5 times. They now have a complete set of my fingerprints. As for facial recognition, really? I just figured those photos were stored on file to show that the customs officer was doing his job.

Re:Great, now commercialize it..

Yeah facial recognition software has been a US Customs thing for a few years now.

"As U.S. airports begin installing face-recognition systems to thwart terrorism in the wake of the Sept. 11 attacks, civil rights activists are rushing to decry the technology as ineffective and invasive."

http://news.com.com/2100-1023-275313.html [com.com]

"In the USA Patriot Act, the National Institute of Standards (NIST) is mandated to measure the accuracy of biometric technologies. In accordance with this legislation, NIST, in cooperation with other Government agencies, conducted the Face Recognition Vendor Test 2002. FRVT 2002 Sponsors and Supporters are: Defense Advanced Research Projects Agency, Department of State, Federal Bureau of Investigation, National Institute of Justice, National Institute of Standards and Technology, Transportation Security Administration, ONDCP Counterdrug Technology Assessment Center, United States Customs Service, Department of Energy, Drug Enforcement Administration, Immigration and Naturalization Service, Secret Service, Technical Support Working Group, Australian Customs, Canadian Passport Office, United Kingdom Biometric Working Group. "

http://www.frvt.org/FRVT2002/default.htm [frvt.org]

I included the long list of agencies because under Homeland Security they will undoubtably share databases. If you have been scanned, everyone has your facial recognition file and fingerprints. I tried to stand out of the camera view, but there was no good way to aviod walking past it. The customs guy did alot of typing when I came in, probably as it was my first time in front of a facial recognition camera. My girlfriend was practically waved through, but she had been though customs just a year ago, as so probably already has a file.

Re:Great, now commercialize it..

Will somebody please explain to me why every time some new way to do anything that would involve identifying people, it's an invasion of privacy? I mean honestly, why are you so afraid?

The freedom of assembly is what's at stake, and it in turn is essential for a free democracy. If the government can track the movements of innocent people, then it can monitor the organizations and associations (including political) that one is associated with. And if the government has the power to document the members of every rival political movement as it is forming, including all the other activities of the members, then they have the power to intimidate and crush it. (Don't believe me? Find a history book.)

Privacy from the government is a key component of freedom, because it places serious constraints on the government's power over the people. Without this, you can very easily become a subject rather than a free citizen.

Re:

Ever heard the term "panopticon"? When you know that someobody could be observing your every move and you don't know exactly when you are being specifically observed, your behavior will change to what you believe the observer wants. Every camera brings u

Re:

When you know that someobody could be observing your every move and you don't know exactly when you are being specifically observed, your behavior will change to what you believe the observer wants.

Only if the observer has the power to impose their will on you. I can stand on a street corner and say I hate the government as loudly as I want, and I don't care if they have a camera watching me. The moment they get the power to lock me up for saying I

Re:Great, now commercialize it..

huh? I said commercialize.. that means, "give me something I can buy" not "give me something for free". wtf?

Surveillance soceity.

I wonder whether these scientists lose any sleep over how their research advances will contribute to the future of our societies.

Re:Surveillance soceity.

I wonder whether these scientists lose any sleep over how their research advances will contribute to the future of our societies.

If scientists ever paused to think for the possibilities of potential abuse of their intellectual effort, progress as we know it would come to a grinding halt. Back to Neanderthal times...

It relies on the ordinary people to safeguard their societies from degenerating but that is an entire different subject (requires getting off the couch alot), and since I can already see the political-zombies approaching to offer their caned insight into the matter it's time for me to split...

Re:

I think you're letting researchers off the hook too easy. There are a million things to research, yet many people choose to work on projects that have dubious implications for society. I mean, sure, there's a lot of gray area between searching for a cancer

Re:

Name five technologies that you like. I bet I can name dubious implications for no less than 3 of them, especially if you make them different, rather than hiding behind fragrances(which can be used to make food of low nutritional value more attractive, or

Re:Surveillance soceity.

Every technology has downsides. This technology clearly has pretty serious upsides. Do you wonder whether the inventors of the integrated circuit lost sleep over the contributions of their work to the surveillance society?

Re:

If something is possible it will be done by someone.

security possibilities

once the technology becomes more refined, they could put a camera connected to a computer(stored with all the mugshots in the nation) near a bank or other commonly visited area be able to catch tons of criminals

Re:

I don't think any of these technologies are up to that level yet, but it's a nice idea.

And yes, the exact same technology could be used to horrific effect by a police state.

Re:security possibilities

Screw the criminals. Thinks about how many law abiding citizens you'll be able to track. :D

Face the Consequences

When a human makes a mistake recognizing a face, they suffer the results. If that's identifying a criminal, they can be cross-examined, or even sued or jailed, depending on what they said that face did.

When computers mis-ID a face, do we cross-examine and maybe punish its programmers?

Re:

Don't be silly.

If this were to be used for criminal identification, I'm sure that when they get a "hit" for a wanted suspect, that they're going to manually sift through the video, in order to figure out direction of travel etc.

These things aren't error pr

No worries...

My tinfoil hat has a visor.

Had to say it...

I, for one, welcome our face-recognizing overlords...

Re:

I, for one, welcome our face-recognizing overlords...

That should read:

I, for one, welcome
HELLO ECTOTHERM.
our face-recog... um. Hello.

My Plugin

if(hot_gurl) {
      RingBell();
    }

A little embarrassed to admit this

improved tenfold since 2002 and a hundredfold since 1995

Because of the internet explosion since 1995, I, too, recognize facials at least a hundred times better. Or at least see them a hundred times more.

Ageing?

How good are computers at recognizing a face after ten or twenty years? I doubt the algorithms can recognize, say, a teenager based on photos taken prior to puberty. Also, can they maintain accuracy even if somebody has a new scar or puts on dark sunglasse

it's in the summary FFS

"for instance, the curves of the eye sockets, nose, and chin, which are where tissue and bone are most apparent and which don't change over time."

Geez.

Re:

They specifically said it uses the visible curvature of the bone as the primary recognition factor. Thinks like scars, glasses, will have a limited effect to disguise.

As for puberty, sure the curvature will change while the skull is still forming, but even

Re:

Really depends what you mean by "headscarf". Some would. Some wouldn't.

Searching google images shows the diversity of things called 'headscarf'. The word covers garments ranging from little more than a headband to the complete head covering a women in Tali

Re:Ageing?

How good are computers at recognizing a face after ten or twenty years?

Hopefully better than I am, otherwise they better be damn good at making generic bullshit small talk at family parties while sweating profusely and fishing for hints without letting on they have absolutely no idea who it is that just ambused them at the buffet table.

Actually, the opposite situation is just as bad. I have enough of a problem with my 'aunt's old room-mate' or equivilant telling me they "remeber me when I was just 'this' big" (given the amount of random old women that at some point 'changed my diaper' I have begun to wonder if my parents rented me out as a training aid), I do not need the computer hardware in my life pulling the same act!

Either way.... this will end badly.

Hmmm...

If my friend, who has no facial hair, puts on a fake mustache, I can still tell that it's my friend.

Will an algorithm be able to distinguish fake markings?

Re:

"for instance, the curves of the eye sockets, nose, and chin, which are where tissue and bone are most apparent and which don't change over time."

Again, it's IN THE SUMMARY.

ORLY?

for instance, the curves of the eye sockets, nose, and chin, which are where tissue and bone are most apparent and which don't change over time.

so, apparently, plastic surgery doesn't exist.

I'm lucky

It's a question of time 'till there 's a law that forbids to wear anything that partially covers your face in certain public areas.

I think I have about ten years 'till computers are able to interpret my front-head as a 'face' so I'm safe.

Since you asked, you can have it...

After LA the incident in May where protesters and cameramen kept running into police batons and shooting themselves with stolen police guns, the LAPD wants the city council to ban masks and goggles [laweekly.com] from public demonstrations. A law somewhere in Europe agai

Re:Since you asked, you can have it...

google can backup any claim

Yep, it can back up any claim.

For example, dinosaurs co-existed with humans [google.com].

Not that impressive

Human face recognition is run by a several hardwired circuits operating in parallel (ie. fast, with little control) with the results put together after by some heuristics -- a good enough guess. What humans need to get from facial recognition, and what their ancestors required and so developed through evolution, is nowhere near the same thing facial recognition software is after. Humans need to recognize quickly that there is a face and what information it's displaying far more than they need to differentiate one from another. Facial recognition software does just the opposite. Also, the software does the complete job every time. Humans only process as much as they need to in any given instance.

If "better" is based on the standards of humans (fastest good enough guess) rather than machines (as correct as possible, complete & in depth), humans win.

Re:

If "better" is based on the standards of humans (fastest good enough guess) rather than machines (as correct as possible, complete & in depth), humans win.

Translation: Throw enough hardware at it, and the machines win? Whatever a computer has been succ

In related news...

Google seems to be deploying something similar [arstechnica.com]

.

Quite impressive..

This is actually a great milestone as we humans are really excellent at face recognition. In fact, we are so good at it that we produce tons of false positives and recognize faces where there are none (clouds, toasts etc).

A few years back (well, nearly a decade actually), I did my master's thesis in a lab that among other things did work on face recognition. The experts there assured me that perhaps in 50 years or so computers might be able to approach human face recognition capabilities. Apparently the development was far quicker than they could have imagined.

An interesting technical point is that in fact the algorithms haven't changed a lot since then - it's still mainly various adaptive systems such as neural networks [wikipedia.org] and support vector machines [wikipedia.org]. The really big breakthrough is in the data collection - in the sensors and scanners. What they couldn't imagine a decade ago was the type of accurate automatic 3d face modeling and measurements that can be done today. It's also how certain computing methods that were deemed unsuitable a few years ago are coming back big time (neural nets for instance). I guess the time wasn't ready for them the last time due to computing power and memory limitations (and of course sensors as in this case).

Caricatures

I'll be impressed when they can recognize caricatures as well as humans.

Other Race Effect

One of the academic research areas I've been involved in, is study of the so-called "Other Race Effect". There is some evidence that people have quantifiable error when asked to identify faces of people of other races than their own.

Computers won't be subject to this.

Re:

And I'm pretty sure the computer won't be able to recognize the face after the facial.

Automating Go

Go is not yet as well-automated as chess, but it appears that go-playing software is rapidly advancing:

"Two Hungarian scientists have now come up with an algorithm that helps computers pick the right move in Go, played by millions around the world, in which players must capture spaces by placing black and white marbles on a board in turn.

"On a nine by nine board we are not far from reaching the level of a professional Go player," said Levente Kocsis at the Hungarian Academy of Sciences' computing lab SZTAKI.

The 19 by 19 board which top players use is still hard for a machine, but the new method is promising because it makes better use of the growing power of computers than earlier Go software."

Link [reuters.com]

See also:

http://zaphod.aml.sztaki.hu/papers/ecml06.pdf [sztaki.hu]

http://citeseer.ist.psu.edu/vanderwerf03solving.ht ml [psu.edu]

http://www.primidi.com/2007/02/26.html [primidi.com]

-kgj