Slashdot Mirror
Facial Recognition Fails in Boston, Too
bryan writes "Only a few weeks after cameras were found to be ineffective in catching criminals in Tampa, FL, a test of a facial-recognition system in Boston's Logan airport also came up disappointing. The cameras which were given photos of employees to detect, were only successful in 153 out of 249 random tests over the past year (about 61%). The article did not say how many false positives the tests generated. The companies involved were Indentix and Visage."
...if the cameras were mounted on black helicopters.
Treehugger? Treehugger... Treehugger!
It has 39% false negatives, not false positives!
Slashdotter are stupid and biased.
I think you have quite a misunderstanding of the difference between "false negative" (which is what the 61% was referring to) and "false positive" (which was not mentioned).
Send your friends messages of love at fuck-you.org
I heard that a very substantial amount of our brain's capacity is devoted to differentiating faces, and it's conjectured that this processing overkill is responsible for such things as people seeing a "face" in the objectively very non-face-like features of the moon.
Give the parallel processing capability people have to do this trick, it's probably not too surprising that computer tech hasn't gotten there yet.
Anyone know more about face-recognition processing in the human brain? I find this topic quite interesting...
~ Whence do you come, slayer of men, or where are you going, conqueror of space?
Watch it turn out the only people it could recognize were wearing pilot hats, or some other highly recognizeable feature like a beard or moustache.
0110100100100000011000010110110100100000011000100
As super-duper as high-tech is, I think even /.'ers would admit that its not a panacea (yet) for all our security ills. The very idea of having a computer capable of accurately identifying one face in thousands -- scanning from afar -- is far fetched. Despite billions in research we've yet to master voice recognition which is, comparatively, much easier to do.
Ah well, what's another few hundred million of tax payer's money shot. I'm sure it made some contractor rich.
CommentBot 0.7a running with args "-module irritate,disagree -target random"
I'm not sure I would call the failure of big-brother tech "disappointing".
"Kelly Shannon, spokeswoman for the State Department's consular affairs office, said the Logan Airport results would not affect plans to use face recognition to enhance passport security"
So it doesn't work, won't help, and might even end up hurting more that a few people, but it's going to enhance passport security?
And Apparently OZ thinks it's a good idea too? "We now have an international standard established, which is the adoption of facial recognition as the international biometric, and that has left us well placed to move to implementation."
Um yeah, we slashdotted usatoday.com, Riight!
wanted: one clever sig,apply within
Seems like the flase positive rate would be the most important stat, and they don't have it.
Obviously it couldn't replace ANY other security measure, but if it worked 61% of the time with NO false positives, I would call that pretty damn successful, especially in such an early implementation.
They said 10 of the 19 hijackers went through Logan - so this system theorhetically would have caught 6 of them? Better than none. And it seems like the technology would improve with time.
Personally I'd rather have my face scanned then have them strip searching me because my credit sucks and I paid cash for my plane ticket.
666-607: 6th floor apartment of the beast
Meir Kahtan of Identix of Minnetonka, Minn., says the technology is far better suited for ''one-to-one'' identification, such as comparing photos on passports or driver's licenses, than random searches of photo databases
It would not take a lot to be 'far better' given the starting point.
They aren't at the stage yet where machines can recognize people based on gait and mannerism. Facial recognition is a best guess and still requires a human to be sure of the fact just like fingerprint systems.
Trolling is a art,
As it happens, a friend of mine is working for a company that are in this field....they successfully implemented eye tracking, (dont tell me, there are lots of companies doing this, but not as well as these guys..). We discussed it the other day, and he told me that the face-recognitions algoritms are coming along..there is some huge stastically problems involved in this, the equipment is not the problem (they are using ordinary webcams) and some special light(ir-freq)...pretty cool stuff. Now, the ethical problems with this are MUCH more difficult.
- While I myself say there is naught, nor ought there be,
Nothing so exalted on the face of God's grey earth
As that Prince of Foods . . . The Muffin!- The Muffin Man
Comment removed based on user account deletion
Indentix and Visage
Is it really any wonder? Those must be the two of the dumbest fucking company names I have ever heard before. Serves 'em right.
Anything about whether employees attempted to disguise their appearance at all. If not I would hate to see the rate if they did.
Either way I don't I like it.
"I think you have quite a misunderstanding of the difference between "false negative" (which is what the 61% was referring to) and "false positive" (which was not mentioned)."
Not misunderstanding at all, just assuming them the same, which is a different crime altogether.
It's probably more like 5% false positive, if it's like any of the other "give us money for antiterrorism" gravy-train projects, which means you only have 500 people per day to strip-search and delay their flight.
Sure, there may be other systems which have a greater success rate, but, at least from what I know (sometimes, admittedly very little), there don't seem to be a whole lot of other alternatives which don't require that the security queues at airports extend out in the airport parking lots. Where no other good solution exists, reducing my chances of crashing into a building by 50% sounds damn good to me.
And, at the end of the day, if you simply force people to take off their hats and sunglasses, so that the camera can get a nice, long look at them in closeup, I wonder how much greater the success rate for facial recognition would be?
I think the concept of facial recognition is being distorted from a tool to help assist in the confirmation of possible criminals and terrorists to a single device that does all the work. The idea that it can still bring in a 61% accuracy rate is pretty good if you compare it to previous technologies. When you combine that with on the ground security and other systems, it only makes it easier for the Airport security, or other government or commerce locations to keep a tight hold on who comes in and out of their systems. Think of it this way, They have facial recognition working at entrances, and a few places along the way to security checkpoints. It picks up 3 positives out of 5000 people for terrorists or criminals (This is just a guess). They send that information to security at checkpoints with a picture from the camera, and whatever might be in a database. The security will be able to check to see if it was accurate at the checkpoint, and make a decision based on that for whether or not to check this person more throughly, stop them, arrest them, etc... whatever the case may be. If they can tell from the photograph that it was probably a false positive, they can just avoid it all together, or pick them for a random security check. No one said if it comes up with a false positive that person will be automatically picked up and thrown in jail. It sounds like a reasonable tool to help identify people they need to check more closely, nothing more.
As with most biometric systems, this is only ever works reliably in a lab.
Remeber the fingerprint system that got fooled by gelatine-gummi's ?
I wonder when these dot-bomb ideas will stop popping back up, and more credible research will get the much needed funds.
There is only one thing that has ever been able to recognize the human face; other humans. (And we do a rather poor job of it too after 10 million years of evolution!!!)
Proof: Take your average ignorant North American, (like myself) and ask him to tell the difference between 3 different Asian individuals. There is a good chance that we would fail that test because we are not used to (or mentally trained to) spot the difference.
{I love using myself for proof, it's so scientific}
"The price good men pay for indifference to public affairs is to be ruled by evil men." ~Plato (427-347 BC)
A government test in 2002 found that face-recognition systems scored correct matches more than 90% of the time when used for such one-to-one identifications.
Once again, the false positives are not given. That is the number that really matters in a society where you can be held in prison indefinitely without a trial or access to a lawyer.
ass-recognition, eh.
I wonder if how my (male) co-workers would compare. It seems like they can sense an ass before it's even in the room. Unfortunately for our national security, their sensors only work on female "terrorists" which are significantly less prevalent as compared to their male terrorist counterparts.
Send your friends messages of love at fuck-you.org
i recognize all the facials in my picture collection !!!
Oddly, I read the article (unusual for a Slashdotter) here and it seems to imply that these companies were marketing their products for the limited use of trying to catch people with forged identity documents. Rather than just having a Customs officer compare the photograph to the face next to that photograph, the software could chime in with "Yeah, that's her alright." It looks like the security people at Logan Airport deployed these products in bulk. I wonder whose bright idea it was to try and use these at randumb? Perhaps a zealous salesperson or an overenthusiastic security manager? I also noticed the company spokesperson sounded a bit "hedged" like the company is trying to state that "Gosh, this product was never meant to be USED the way this customer is using it." The part left unsaid by the spokesperson was, "We told them this wouldn't work..." On a side note, let's not even consider how abysmal this software must perform when terrorists are deliberately disguising their faces.
In principio erat Verbum.
Sounds impressive, actually. If there are 4-5 hijackers, and each has a 61% chance of being noticed, then the odds are good that at least one will be and the plot will be foiled.
Also, what's the worry about false positives? If and when they happen, it's a simple matter to clear up a person's real identity. It's not like they shoot first and ask questions later.
They said 10 of the 19 hijackers went through Logan - so this system theorhetically would have caught 6 of them? Better than none
The 9/11 hijackers used their real names and real ID. If they'd been placed on a simple watch list of names then strcmp would have found them, not some highfalutin' face recognition system. It's not the technology here, but coordination between the three letter agencies that's needed.
John.
I'm not sure it's even worth looking at the data if we can't have some idea of the level of false positives. If you can find 1 out of 1000 criminals that walk by seems that it might be worth it?
What you don't want is harassing innocent people. If we can aviod that, I don't see where the problem is.
Have you kicked your kitten lately?
Oh dear...
First, a poster of someone else's face (facial recognition evasion).
Second, the goey fingerprint duplicator,
now this walk-by signature hacker on a PDA?
What would be next?
Hijacking IRIS pattern (simply stareing at the bathroom mirror)?
Stolen DNA pattern?
There is no solid defense against unrevokable but stolen biometric parameters.
Assuming that the remaining 39% were false-negatives, then I think that the system worked incredibly well.
Think about it in terms of spam. No one solution will stop 100% of the spam destined for your e-mail address. It takes a combination of methods (and even then, you can only approach 100%, never achieve it).
The same attitude should be taken in airports. A system should not be dropped because it's not 100% effective. It should be used to strenghten existing and future security.
I wonder why the inaccuracy of this system wasn't well known before it was put up in a public place. Did it perform much better under the controlled environment of the lab? The article states that it works well in a one-to-one test, but they knew that this isn't how how it would be used in this case. It seems likely that if this failed so miserably in real life it couldn't have been that great when they were developing it. Does this speak of a certain desperation on the part of law enforcement to 'do something' or at least to appear to be doing something. Or maybe a hopefullness on the part of the company developing it that they might just get lucky. In fact, if they were payed by the government to deploy this test even though it seems likely they knew it would fail, maybe they did get lucky. Who payed for all this anyway?
How exactly does the fact that someone is gay make it harder or easier to recognize them?
When the choice of camera used by some systems have quite a low resolution.
You need very high quality images for recognition to work well. Try OCR-ing a badly skewed very low resolution scan and that's just text.
With facial recognition you have to worry about shadows, different angles, glasses, changing hairstyles, facial hair and so on....
Wouldn't a "false negative" be a "positive"?
"They" are just people like you and me. You won't be held for any length of time without good reason, and the fact is that you're completely wrong about holding people without charging them.
"{I love using myself for proof, it's so scientific}"
Hehehe, but you do raise a valid point. If we, the programmers cannot really do the task how do we get a computer todo it reliably.
Don't forget that there is more to ID'ing someone than just what they look like. Like speech context has everything todo with it.
You're in the middle of a conversation and one word is muddled. Your brain can quickly come up with probable words in a manner of a few seconds. This is all based on context.
Similarly you catch a glimpse of someone that looks like your co-worker *and* you're at work chances are your brain will make the ID much more faithfully then say you were at a Mets game or something...
Tom
Someday, I'll have a real sig.
Okay, I admit it. I single-handedly foiled big brother's plan by marching around Logan with novelty glasses and a giant foam cowboy hat.
Not quite. It would only have caught those of the hijackers who were on the do-not-fly list, which we all know is a resounding success. Since the hijackers did nothing to arouse suspicion in their initial period in the US, it's unlikely they'd have been flagged.
Who comes up with harebrained schemes like this anyway?
It sounds like something a couple of potheads thought up.
Engr1: Dude, you know what'd be awesome? We could make a widget that recognizes faces, then we could put it at the door so we'd know if it was the pizza guy knocking.
Engr1: Whoaaa dude, that'd be awesome. Pass the caffeine.
The current state of the art in facial recognition doesn't allow for automated operation. Anyone familiar with the technology realizes this. Articles such as this are no great revelation to anyone with a clue.
The current tech certainly is imperfect. However, it does offer the potential to allow increased scrutiny of a subset of passengers. This alone has value. We just have to decide how many false positives we can manage.
Just contemplate if 5 of the 10 hijackers at Logan had been detected. Every little bit can help.
A Pirate and a Puritan look the same on a balance sheet.
Besides, "limited" access is not the same as "no access," now is it?
If a false positive indicates someone who was incorrectly identified as a positive match, then a false negative would surely mean someone who was incorrectly identified as a negative match. Brian K
Unforutnately the first is not a valid substitution option for the second. As someone stated above(below, whereever), not ALL terrorists are in the DB's, so scanning the face of wannabe(or will be) terrorist will bring exactly the same result as scannig the face of Marta Stewart....well, except that she'll probably ring a bell and be get back in jail;o)). If the visual recognition technic is to be adopted it will provide another back-up for the strip searching not a substitute.
1. No sig. 2. ???? 3. Profit!!!
..hugh amounts of tax money will fix these little problems.
Thank you.
"You can't bolster an already misinformed argument by simply making up statistics on the spot like that."
If you mean by "misinformed" something other than "having read much of the technical literature available", you're wrong. I have.
As to the "making up statistics", it was admittedly hasty to post from memory rather than looking it up. Actual tests show 0.3% false positive in "controlled" (i.e. setup) situations, and 10% false positive in real situations. So, twice as bad as in the post you responded to.
Do you want to continue haggling statistics all night? Even if some hypothetical system were 99.999% accurate, you'd still have a queue of terrorists at each airport terminal.
http://www.theregister.co.uk/content/4/21916.html
Do you have a well-lit, correctly exposed photograph of Osama Bin-Laden, taken from less than 2 meters away in controlled conditions? How about his aides?
If facial recognition technology is used alone to try to catch criminals, of course it will fail. However, when used in conjunction with other systems, even a low success rate can be helpful.
I have no idea how the system works, but my guess is that it was spitting out "definitely yes" or "definitely no" when it should instead be offering an estimate. "73% chance that this guy is the same person as this felon who is on the lam".
It's not the kind of thing that stands up in court, but it is the kind of thing that can raise an eyebrow, and in those cases, that's often enough.
That being said, the test itself doesn't sound very good. It's quite easy to fake tests of software, especially if you have a limited pool of data, or know what you're looking for ahead of time. It's possible the result would be even worse in actual usage.
"The Sage treasures Unity and measures all things by it" - Lao Tzu
Maybe they could come up with some other recognition to use in combination with the strip search? :-)
666-607: 6th floor apartment of the beast
I haven't followed facial recognition too much but...wouldn't twins have an issue with this?
I'm an identical twin, I've had my lights punched out by someone who thought I was him (Thanks bro..grumble)...
Anyone able to tell me how this would differentiate between siblings that look very very much alike?
It quotes the Logan report saying, "the number of system-generated false positives was excessive."
Thank god.
I'll be happy not to have the computers tracking my face thank you very much. Too much of a pkdick idea I think.
But what about the system installed in London, England?
simon
home page
That someone who was frequently incorrectly flagged by these devices would become quite adept in proving his innocence in a timely manner.
They'd have been cleared and let on the flight, dumbass. They used thier real names and IDs. This technology wouldn't have prevented the Sept. 11th attacks, but may, with another 5 years or so of improvement, be of use in preventing other attacks with different methodology.
But what about the system installed in London, England?
That one doesn't track faces per se. It indexes along bad teeth.
Comparing it to Windows will be a moot point, since El Dorado is going to have a 40% larger code base than XP.
However, it may be useful while using it with a passport picture. In fact, last week, canadian government advise passeport requesters they will be refused if the picture doesn't conform to a new requirement. It is strictly forbidden to smile on your passeport picture. I don't know the name of the faces recognition software they use, but it is likely it may be completely by pass by too much happiness!
Is a kamikaze-terrorist's smiling before pushing the little red button?
Achille Talon
Hop!
Just thought I'd chime in here. Technology is neutral. It can be used for good or ill. Facial recognition technology can be a great thing, if used properly for constructive purposes. For example, it could be used to help with identity recognition at ATM-machines.
Yes, these technologies are failing alot. But, just a couple of years ago, people would have scoffed at the idea that computers could even begin to accomplish some kind of face-recognition. This technology is in it's infant stages. I don't think you can blame a technology that's just gotten off the ground for not being perfect.
Lets criticize improper uses of this technology, not the technology itself.
social sciences can never use experience to verify their statemen
it means 31% of the employees weren't recognized as employees. It should have been a positive match, but returned a negative, falsely.
Yes! Mod this up! These systems aren't meant to replace the security personnel already in the airport. It's simply a tool to make their jobs a little easier...
Funny how the tinfoil-hat crowd automatically assumes the worst when something like this pops up.
Although i find the technology fascinating, and when put to good uses it can be very useful....
I know there's plenty of people out there ready to exploit it or exploit the public with it.
So in a way, i'm kinda glad it doesn't work as hoped...
do() || do_not();
The most important characteristic of such a system is the false positive rate. A system that flags everyone who passes through will flag 100 percent of terrorists but would be no better than having no system at all. They do not give the false positive rate but it is highly unlikely to be less than 10 percent and may be much larger. Since the ratio of terrorists to non-terrorists is probably on the order of a billion to one a system with an unrealistically low 1 percent false positive rate will flag 10 million non-terrorists for every 0.5 terrorists if it has a 50 percent correct ID rate. Even if you do extra searches on those 10 million people, with a 50 percent correct ID rate the terrorist is just a likely to be in the 990 million people who do not get flagged as in the flagged group.
You need a close to 100 percent correct ID rate and a false positive rate below one in a million, which is probably impossible, before the system would be of any use. However all this assumes that you have pictures of all terrorists. This is just plain impossible, especially in the case of suicide attacks. This is not like bank robbers where there are multiple incidents allowing evidense from witnesses etc. to be used to catch them when they try again. With suicide attack the attackers will likely be model citizens (who will not be on any list) right up until the attack and afterwards any info on them that is gathered is close to useless.
Remeber the fingerprint system that got fooled by gelatine-gummi's ?
/d/ and /dh/, even with fairly extensive training. A computer can, however. I imagine that a computer would also be at least comparably able to tell apart similar-looking Asians.
Yeah. Thats' what guards are for. While they might not catch your gummy finger, they'd probably notice you holding up a George Bush mask as you walk by the scanner.
Careful disguises are a little tougher, but from what I've read about these systems, they do better than your average human against disguises (they use different clues, and don't put as much weight on obvious things like facial hair). This might put your human past the tolerance though, as you'd have to set tolerance pretty low to avoid excessive false positives.
Proof: Take your average ignorant North American, (like myself) and ask him to tell the difference between 3 different Asian individuals. There is a good chance that we would fail that test because we are not used to (or mentally trained to) spot the difference.
Fine, but there are things that the computer can tell better than humans, even things that humans "should" be able to do better. For instance, your average ignorant North American can't tell the difference between the Hindi consonants
I hereby place the above post in the public domain.
They held him as a "material witness." He was allowed to see his lawyer and have contact with his family, contrary to what the other poster said. There is absolutely no evidence presented thus far that the FBI held him as a material witness in bad faith. The fact that he was later charged with a crime is irrelevant to this concern. The material witness statute isn't exactly new, anyway.
The system can be used to recognize a particular face when it is standing alone. Consider, for example, a photo of a face sent along with an visa application to the American embassy. Please read "World: Asia-Pacific China backs embassy protests". In 1999, Serbians committed gross human-rights violations against the Kosovars in Kosovo; the Chinese fully supported the Serbians in their campaign of terror. The North Atlantic Treaty Organization (NATO) under American command attempted to stop the slaughter by knocking out Serbian military units. NATO deliberately attempted to avoid hitting civilian targets in Serbia, but some bombs accidentally hit the Chinese embassy.
Shortly thereafter, the Chinese in both China and outside China erupted into ugly, violent protests. The Chinese throw stones and other projectiles at the American embassy in China. The Chinese also attacked some Americans. " The residence of the US Consul General in the south-western city of Chengdu was stormed and partially burned ."
How could the Americans in China have responded to this nonsensical violence? The Americans should have done the following.
-
Pull out cameras and take pictures of all the protestors.
-
Scan the photos into a computer and transmit them to Washington.
-
Henceforth, when a Chinese submits an application for
a visa to travel to the USA, use the face-recognition system
to determine whether the photo of the applicant matches any of the
protestors. If there is a match, then the application will not be approved.
- At the American embassy, grab a megaphone
and loudly announce, "Attention protesters. We are using
a face-recognition system. Any protestor applying for a visa to
the USA will be denied entry into the USA. We are taking pictures
right now. "
After about 10 clicks of the shutter of the camera, all the protestors would have disappeared. Henceforce, we should use this face-recognition system in conjunction with photographic equipment at all embassies and consulates run by Western nations within China (which includes Taiwan and Hong Kong).The computer simply does not recognize members of the "attractively challenged" brigade...
"Right now, somewhere in this world, Scott Baio is plowing a woman he doesn't love," - Peter Griffin, *Family Guy*
Hey, dumbass, you could take box cutters on planes at that time.
Is an assumption made in the courts, not in the field. Otherwise, how could police officers point their weapons at suspected bank robbers? This is common sense.
Seriously, I know a guy who moonlights for the pentagon, he's one of the inventors of 'digital video storage', AFAIK.
:-)
Every few weeks a few burly men in a caravan of Ford SUVs pulls up and he's off for a few days. He just bought a house in southern france, 9/11 didn't hurt HIS economy
Seriously though, great guy, It was an absolute joy to live next door to him.
"Sometimes, I think Trent just needs a cup of hot chocolate and a blankie." -Tori Amos on Nine Inch Nails
Cunningly facist, I like it.
Someone can be a known suspect and not be in custody -- ever seen America's Most Wanted?
Since I don't have access to the court records it's tough to know anything, but it can safely be assumed that there was good reason as the man plead guilty to a crime.
I dislike the notion of being watched, categorized, and monitored everywhere I go. At this point, facial recognition systems have proven to be relatively inaccurate, and thus they have failed to gain widespread acceptance.
Proposals for facial recognition systems continue to be shot down because of their inaccuracy, but why does it have to be their inaccuracy that is the sticky point. Shouldn't the fact that they constitute a massive invasion of privacy be all the argument we need?
If we continue to use the "accuracy" argument over and over, then what happens when a system that is proven to be fully accurate comes out?
Facial Recognition Systems aren't a bad idea becuase they're inaccurate, they're simply a bad idea -- and that is what we should focus on.
You miss the point of the article. This was a small scale test where only about 140? individuals data was entered.
If you ramp this up to a full system then the following happens:
1. The biometric data of MILLIONS of people has to be stored. This exceeds current capacities of available computers.
2. Response time will SUCK worse than current x-ray machine lines.
3. The false positive and false negative rates will soar with the addition of the other data to compare to.
Lets face it, this test failed MISERABLY. It is time to take our tax dollars elsewhere.
Actually strcmp would have problems - real names tend to be properly capitalized while ID's tend to have names in all upper case. So the better string compare function to use is stricmp. Then there would be problems with leading/trailing spaces - inept ticket counter employees, you know. So run the names through ltrim( rtrim( ) ) first...
Prisoner of War is a status given to legitimate combatants under the Geneva convention.
Other combatants are war criminals, who are specifically not given rights under the same convention.
As for "our" freedoms, speak for yourself. The government is given discretion without due process over international affairs to protect those very freedoms you're so concerned about and it has not yet been demonstrated to act in bad faith in that persuit.
If your country allows the government to detain you indefinitely without arrest or communication or access to a lawyer (under the guise of anti-terrorism or homeland security) , you should be VERY concerned about false positives.
The US doesn't, and even if it did, it would double-check things beforehand.
But I'd like to see what would happen if these systems designed to recognize faces from afar in public streets and parks started registering thousands and thousands of sightings of known terrorists on the monitors down in the Big Brother Bunker.
And there are humans looking at these face systems. They'd probably pull up the tapes, figure out you were being a jackass, and get on with things. Even if not, what would they do if you suddenly ordered several truckloads of ammunition and ammonium nitrate to an urban address? They'd damn well watch you, because there aren't that many jerks who go out and impersonate terrorists for kicks.
Call it monkeywrenching. Call it free expression. Call it a costume party.
This is for airports. There are guards. They are not (all) blind. If you try to fool the cameras, you will be searched, and free speech does not matter. If you make yourself a false positive, it's you that suffers.
Now the tech people may scream "those aren't false positives! the system is correclty identifying Mr. Terrorist!"
If you're using this to decide whom to search (instead of to track someone), this sort fo false positive doesn't matter. If you put 10 jars of peanut butter in your luggage (looks like plastic explosives to scanners) and get searched as a result, this also is a false positive. (More worrisome is breaking the jars of peanut butter and smearing them in your bag, along with some well-hidden plastique, but this attack probably wouldn't work and does not generalize to face scanners anyway.) If you go around talking about how best to blow up a plane, you will be kicked off the flight; that is also a false positive. Nobody complains about this. Nobody cares if the airlines inconvenience jackasses who intentionally trigger their security system by impersonating hijackers.
I hereby place the above post in the public domain.
The Federal Governemnt enjoys many, many powers not given to it explicitly (or even intended for it) in the Constitution. These are granted to it by interpretation of the document by the courts. This is just how it works and always has.
The number of shaky premises, false inferences, and unjustified conclusions in your post, is right up there with mainstream journalism.
Rather than show all these problems in formal logic, let me just point out one problem:
A humans ability to do facial recognition has nothing to do with a computer's inability to do it.
Also let me throw in the technical point that any parallel solution to the problem, can be emulated serially; it just will take more time and probably more hardware to do it.
I think the slashdot crowd are going about this all wrong, geeks just cant see past the technical side. Ok, if your a large government wose citizens are worried about a potential repeat of the largest terrorist attack in your history then what are you gonna do? Sure, you could make big solid security changes, but that costs money, and your main objective is to provide peice of mind to your citizens for, well however long you plan to stay elected. Now you know its unlikely there will be a terrorist attack, atleast until your long gone, but you can win votes and confidence by throwing sums of money at various band-wagons. Face recognition is the perfect one - you and i know its next to useless in its current form, but the general public know that they've seen it in sci-fi so its gotta be good right?
A company can develope it (read dilbert to see a good idea of what that company might look like) and they can get monet from the government or the airports or whatever:
1) Invent technology that might solve current problem that everyones facing
2) ?
3) Profit
This comment does not represent the views or opinions of the user.
I can't find right to fly in my constutition anyplace? Anyone else miss this?
As you can see I don't care about my karma.
Phillip K. Dick would love this discussion.
Here we are discussing the government using
computers to classify people as terrorists, to be
treated accordingly, like it was a reasonable thing to do.
The vicomp classified humans versus androids, who
where then put to sleep - even in the case of false positives.
TRUST THE GOVERNMENT - THEY ARE HERE TO HELP.
I rather suspect that the false positive rate would make this system impractical. If there were only one false identification per ten thousand people, then the false positives would significantly outnumber the true identifications. This has been pointed out to be the real weakness in biometric identification systems.
[nt]
Even if facial recognition was 100% accurate, it still would not have prevented any of the problems with airport security. For some reason, all of the "new" security measures at airports (no nail clippers?) really don't address the problem. Remember, the bad guys all had valid ID's. No facial recognition needed. There names were printed out on thier tickets, ID's, heck, they were printed on the aircraft manifest.
The only technology that could have helped would have been a mechanism for law enforcement to flag these folks, but that would require government to actually provide usefull services. Oh, and not to mention the whole invasion of privacy that Mr. Ashcroft is shooting for.
--WooooHoooo--
Of course, the 61% recognized were all Kennedys.
Taken from the ACLU web site:
According to the Logan report, which was written by an independent security contractor, "the number of system-generated false positives was excessive, and as a result, the operator's workload is taxing and strenuous, requiring constant undivided attention and periodic relief, which amounts to a staffing minimum of two persons for one workstation."
blah blah blah... the only thing I'm interested in knowing is *WHAT* causes the recognition failures? It's not mentioned in the article. People aren't discussing it. BUT it's a very important question... especially for those concerned with retaining private identity.
So, can it be assumed that wearing sunglasses will defeat the system? What about keeping your face downward to the floor? What else? How about walking through the airport terminal with a cup of coffee held up to your mouth? If that's all it takes, I'm wearing sunglasses to the airport from now on. It's that simple. In fact, I'll wear sunglasses at any venue where facial recognition "technology" has a chance of being used.
Skiers and Riders -- http://www.snowjournal.com
Seems like the flase positive rate would be the most important stat, and they don't have it.
Obviously it couldn't replace ANY other security measure, but if it worked 61% of the time with NO false positives, I would call that pretty damn successful, especially in such an early implementation.
Uhhh, what? So, say in the case of a terrorist. You'd rather this scenario : 40% of the terrorists get through without being recognized and possibly cause the death of thousands of people, but anyone who's NOT a terrorist doesn't get five minutes of inconvenience. The alternative scenario, of course : a couple innocents get searched unecessarily, but no terrorists get on the plane and kill people.
To me the latter sounds like a more successful implementation.
With quantum computer problems like this will be a thing of the past. We will just say "locate all criminals" and then the quantum computer will do it because they know things without actually knowing them. We will tell them to find the crimincals but they will have already found them, before the criminals even knew they were coming. Quantum computers will also be good for telling us what TV to watch and then rather than just watching the TV it will tell us whether the show we were going to watch was any good. Saves time that way. And then they never really had to make the show because the quantum computer already knew if we were going to like it so there is no real point in actually spending money to make the show. At least, this is my understanding of how quantum computers work. I think you can also do floating point math up to 10 digits too, but I wouldn't hold my breath.
Similarly you catch a glimpse of someone that looks like your co-worker *and* you're at work chances are your brain will make the ID much more faithfully then say you were at a Mets game or something...
It'd actually be easier to catch a terrorist at a Mets game than at an airport - at a mets game, you only have to pick him out of about 1000 people!
*rimshot*
It's well-documented that there was good reason to believe that Iraq had WMD before the invasion; the fact that none has yet been found doesn't prove that the leadership acted in bad faith. See President Clinton's speech concerning Operation Desert Fox launched against Iraq in Dec. 1998. -- President Bill Clinton, Dec. 16 1998.
Four years passed with no weapons inspections. That's forty-eight months.
That means the technology isn't ready for prime-time yet, in any significant use. But the only way to get accurate data is to test it out (not necessarily to rely on it, but to test it against more traditional means that we know work).
It's sort of like voice-recognition in it's early days.
social sciences can never use experience to verify their statemen
With no mention of false-positives, 60% recognition is exceedingly accurate. Nobody's talking about using this technology by itself, but imagine if 60% of criminals who would have otherwise gotten away were able to be aprehended- that is not failure.
this is only, of course, with no mention of false positives.
-- 'The' Lord and Master Bitman On High, Master Of All
On the other hand, Beowulf clusters are pretty reasonably priced these days.
However, you are correct in that the processes and structures the human brain evolved to perform pattern recognition (of which facial recognition is just a small subset) are of limited consequence when it comes to machine vision. Not that we really understand how the brain does what it does, yet. Eventually we will, but I suspect that by that time we will have perfected machine vision to the point that we won't care.
The higher the technology, the sharper that two-edged sword.
They made a decision that you don't agree with for political reasons so you'll discount them entirely. Idiot.
There seems to be 2 different ways to us this technology: 1) As verification in the scenario of someone presenting an id and a camera pulling a current image and the system verifies that the person presenting the id is the person on the id. I think the jury is still out on how effective this will be. 2) As recognition in the scenario of mass amounts of people moving by and the system identifiying someone as a terrorist and sending the authorities to nab them. I doubt this will ever work. Hell, humans can't do it because of disguises and we have massively parallel systems between our ears. How many times is a recognition system going to get a good enough image capture of a person to enable a certain match against a criminal/terrorist image database? What happens when someone on the list tacks on a fake mole or scar? Granted the software has to be able to take some variances into account, but at some point giving it that ability also gives it the ability to generate a high number of false-positives and become absolutely worthless. Is this too simplistic view of the recognition system?
TLHO
Twin or more? ITA
Apache/Spring/La
You can narrow down the false positive stat somewhat by realizing that ff it correctly identified 61% of the people as security risks, it should statisitically *IN*correctly identify as much as 61% of the people who pose no security risk at all. The only way it could misidentify more than 61% of innocent people is if the software is abysmally broken (in which case, the fact that it correctly identified 61% of non-innocents is utterly miraculous, and likely attributable to coicidence).
File under 'M' for 'Manic ranting'
It succeeded in earning the makers of the hardware/software bundles of cash.
Can't be detained without a warrant, which requires, you guessed it, probable cause. A person held as a material witness has a right to counsel and his being held must be approved by a judge in a hearing.
I think you missed the point. If you're the programmer of the system and you cannot see the difference between images of three people, how on the earth are you going to come up with an algorithm for the computer to do the thing instead? You can try something like neural networks but unless you understand what to look for, the changes are pretty slim that the system works at acceptable quality. I consider programming as writing formal instructions to do something - how do you teach something you've no faintest idea about?
Where's the research results for this area? I guess inputting different measurements of the face for many enough test subjects, vectorizing the data set and throwing off the axles that contain very little information could provide some generic information what to look for.
_________________________
Spelling and grammar mistakes left as an exercise for the reader.
Yeah, it was a bad idea to begin with, but let's give Boston and Tampa (Bay?) some credit.
The government saw once of its "law enforcement" / "war on terrorism" programs was ineffective and (gasp!) dropped it.
Isn't this exactly what we're *not* supposed to see from this bloated, non-responsive, heavy-handed bureaucracy/police-state that the libertarians/progressives bash all the time?
They had a trial run of a new, controversial idea and it didn't work. Isn't that exactly the sort of innovation and creativity people claim "big government" sorely lacks?
Joe
http://www.joegrossberg.com
a spade a spade. It saves time.
It's a plain and simple tactic. Privacy, surveillance and big brother related news stories have really picked up in the last few years.
Disinformation.
Plain and simple. I don't know if you notice in your town or on your freeways that more and more cameras are being installed. Take a look. Oh sure, we turned off the facial recognition technology..wink wink.
90% of the intersections in my town have 4 video surveillance cameras, these are NOT red light cameras. We only have anout 3 red light cameras. The cameras are not installed on state highways, in city limits, in my town.
THIS HAS BEEN A SPECIAL NOTE FOR THE CONSPIRACISTS.
The Bush Administration did not claim that it had "incontrovertable proof" of the weapon's existance. I'd ask you to produce a quote to support that statement, but I know that you can't.
And also, what part of "last chance" (as Pres. Clinton threatened) don't you understand? After so many "last chances," someone had to say that enough was enough and that was President Bush. I commend him for it.
I've taken the libery of fixing it. Now we can finally replace those hard-working security screeners:No thanks needed; I'm just glad I could help my country.
It seems to me that face recognition would have to be just about the worst thing to try to identify people by programatically.
I'm sure it's all the name of mass scanning the public from an already compiled database, but it's such a difficult metric to attemt to track people by. Besides the obvious failure rate with the method, surgery and facial trauma can gravely affect the outcome of the scans.
If you're trying to check face proportions against a huge database of pictures, you're going to have mass failures. (false negative and positive) It's hard to have labratory conditions at an airport and proportion differences can be very slight.
I'd rather seeing mandatory Iris or Retinal scans at all border points and identification procurements.
You get to a point without your bio in the database, expect to be scruitinized.
To me the latter sounds like a more successful implementation.
What if you get a figure like %99.99 of all alerts are false positives. At that point, human nature causes the security personnel to ignore it entirely. The problem with 'alerts' that go off all the time when nothing is wrong is that we naturally come to see them as a 'normal condition' in much the same way you don't really hear the fans in your PC after a while.
Apparently, our Government (U.S.) is determined to de everything BUT the sensible thing.
Forget the extra screening, no fly rules, strip searches, multi million dollar machines that don't actually work (or even go ping), etc and DEADBOLT THE COCKPIT DOOR. Ace hardware has all we need to secure the airlines. If that doesn't feel like enough, place skymarshalls on random flights, dress them as business people and tourists.
I love these articles. Leaves me wondering in which exotic location this technology will fail next... 8)
Diplomacy is the art of saying, "Nice doggie!" until you can find a rock.
As someone who's successfully made a facial recognition system, I can say it's not far fetched at all, nor is it something "20 years from now". The tech is all there, it's just a matter of getting it to talk, and applying it correctly.
We can make systems to identify faces, all by themselves, with nearly 100% accuracy and no false positives. The limit of such a system is the volume of the sample material. This is esepcially true with fixed or predictable backgrounds. Likewise, we can make software to find and isolate faces in a picture. I've seen it done.
It's not improbable, it's just not ready yet.
Slashdot. It's Not For Common Sense
Recently there was a story in the press about a new rule for Canadian passorts: no smiles. Now we know why...to normalize the faces for facial recognition.
To say that his claims are lies is to say that he did not genuinely believe in the reliability of his sources. You can't even prove that he didn't have good reason to believe them.
But far too obvious. Try again, loser.
You make the assumption that everyone in the world harbours a desire to go to the US.
Trust me, this is not the case.
Reminds me of the old Saturday Night Live skit of drug inspectors at an airport. They spent all their time inspecting "suspicious" people like Garret Morris, but let priests with caseloads of cocoaine walk through.
I've found that my posts don't format quite right w/o a sig.
My dog does pretty well.
I've found that my posts don't format quite right w/o a sig.
I live in Boston and hardly anybody recognizes my face.
These kinds of systems are measured by false positive and false negative rates. A 40% false negative rate with a 0.01% false positive rate would be quite useful for law enforcement: you catch more than half the fugitives walking past the camera (when currently you catch none), and you only inconvenience 1:10000 travelers. A 40% false negative rate with a 10% false positive rate would be unusable because security would spend way too much time trying to check whether every 10th traveler is a fugitive.
One speech. I won't debate the merits of the claim, but it was hardly the only piece of evidence considered.
I don't know your face, but your name is familiar.
This tagline is copyrighted material. Please send $10 for an affordable replacement.
No, actually what I meant by "it couldn't replace ANY other security measure" is that if it could hit on 60% of the 'bad guys' it sees, that's better than 0% of not having it at all.
It's one thing to get searched "unecessarily" once in a while, but like in the case of people who happen to have the same name as someone on a watch list, it's every single time. If the mathematical "score" of your face structure rings a bell, it's going to do it every time. Can you imagine if you had to fly 3 or 4 times a week, and had to endure 2 or 3 hours of questioning to prove you're not that John Doe? And there's nothing you can do about it. There's even one case (sorry for the lack of links - I heard the story weeks or more ago on NPR) where a guy is getting hassled every time he flies, even though the guy with the same name that that's on the list has been in custody at Gitmo for months.
I'm all for a little nusance to save lives, but some of it is way overboard. What would have prevented 9/11 has less to do with airport security, and more to do with better management of info. The info was there - there were reports that they were planning to hijack planes and use them for an attack on US soil before 9/11. Funny you don't hear a lot about that, huh?
666-607: 6th floor apartment of the beast
Actually no. Arabic names can be transliterated many different ways into English.
I think you missed the point. If you're the programmer of the system and you cannot see the difference between images of three people, how on the earth are you going to come up with an algorithm for the computer to do the thing instead?
First of all, it is often easier for a human to tell the difference between detailed images of three people than it is to tell between the people IRL, since images can be studied in detail and at length. The problem is generally that the distinguishing features are things that you don't normally notice when looking at a face, because of looking at only one race. In this case, I think that programmers knowing what to look for in Asian faces is not a problem.
The computer has the advantage for both of these reasons. Often, general methods of solving a classification problem work on variations of that problem, even if specific ones don't. In text analysis, for instance, Bayesian and chi^2 statistical filters, and whatever generalization of Bayes' thm CRM114 uses, can classify other sorts of text; they can tell, say, philosophy from history essays. I expect that in facial analysis, a general method which can tell apart two Caucasians accurately would work on two Arabs.
In face and sound recognition, humans are somewhat crippled in that information fed to recognition systems is temporary (it doesn't all stay in memory as the person moves), and it has been extracted by other subsystems from much more detailed data. While these subsystems are very effective, they have been trained on a very limited set of data (people/voices you know), and so may not be effective on data outside that range. With a computer, you have to develop entirely different methods, and train them on a wide variety of people. While this is extremely difficult, it leads to a more general algorithm.
I don't have any research results for face recognition, but I have heard about speech recognition. Many current speech recognition systems today are language-specific, because they are based on clues derived from speakers of that language. But my roommate last year worked on a principle-component-analysis system (more or less those axes you were talking about) which could (in theory) be trained to work on basically any language. The training did not need to be to an individual, but it was very processor-intensive, and recognition wasn't cheap either. I don't think he developed it very far, though (as in, phonemes->words, and optimizations to reduce the CPU time), and ended up applying the technology to voice compression. His compression tech provides very good compression, possibly the best ratio in the field for a given (subjective) quality, but also requires a buttload of CPU. Therefore, it can't (yet?) be used in the intended market, which was cellphones. He estimated that a 600MHz PC would be required to compress and decompress in realtime, so this might be useful for VoIP-type technology.
I hereby place the above post in the public domain.
Anyhow this is what happens when you have a bunch of programmers from India working on your code.
and just think, pretty soon these guys in India will have your job!
Spread the RC luvin'
Anyone who spends five minutes thinking about the implications of a 5-10 percent error rate could only reach one conclusion about these systems. Now those multi-million dollar baggage scanning machines that they *had to have* are next you know. It's all a huge waste of time and money because you can't use a system where the error rate is higher than the rate of possible success. The TSA is a monumental boondoggle.
Technically speaking then, as we have pictures of Saddam Hussein and Osama Bin Laden, they shouldn't be able to apply for visa's to the USA?
Doesn't plastic surgery stop this project in it's tracks, or maybe growing a beard or getting a haircut? I heard the AU Customs want to do something similar to this.
How 'bout that robot made by Honda Corporation, it recognises faces, and even waves too. Maybe they could ask Honda to lease it out for scanning Visa applications?? At least it's friendly and can probably be programmed to make coffee!
[ess]
Taking links from a website I know cannot handle a Slashdotting:
Waleed M Alshehri - alive and well in Casablanca, Morocco
Marwan Al Shehhi - Alive; same link as above
Ahmed Alghamdi - Alive; same link as above
Wail M Alshehri - Alive
Ahmed Alnami - Alive; same link as above
Abdulaziz Alomari - Working for Saudi Telecom
Khalid Almihdar - alive and living in Jeddah, Saudi Arabia
Salem Alhamzi - Working at a petrochemical company
Saeed Alghamdi - Alive; same link as above
We do not know who the hijackers actually were.
I thought I had fully qualified my statement with "I heard" and "it's conjectured", but go right ahead and enumerate the shaky premises, false inferences, and unjustified conclusions you can extract from a 2-line opening of a topic. That's what I posted it for, as I would hope the last line makes manifestly clear.
There are many, many examples of things humans can do in the context of their "computing architecture" that rely on the biological equivalent of "parallel processing". Vision is one of them, catching a ball real-time is another.
"Nothing to do with..." So what? It's an analogy. If the topic was nuclear fusion, and I drew an analogy to the sun, would you then respond with "The challenges of nuclear power plants have nothing to do with a star" or some equally pointless, uninformative response? If you are trying to design a system that poses seemingly unresolvable problems, looking at another system of *any* origin that *does* work could be considered an appropriate course of action, to generate ideas if nothing else.
Obviously any parallel solution can be performed serially, given time. The context of our evolution, however, has pressed upon us the need to generate real-time responses to large amounts of sensory data, which is exactly what's needed for the original story's requirements. But thanks for the wholly gratuitous, self-evident information.
Please at least make your next troll a little less content-free.
~ Whence do you come, slayer of men, or where are you going, conqueror of space?
Well, I agree, but not on lack of programming ability. I think it's a lack of imagination, or a fear of trying new technologies. I have friend who runs a game dev shop out of hong kong. He says he loves it because he can get ten programmers (with equivalent skills to an american programmer) for what he would pay two programmers in the states. The problem he has experienced repeatedly though, is that they (the hong-kong-ese programmers) follow directions exactly, but lack any imagination whatsoever. They do exactly what they are told, and no more.
Spread the RC luvin'
I wonder when these dot-bomb ideas will stop popping back up, and more credible research will get the much needed funds.
When government spending is allocated based on merit rather than scratching the backs of constituent contractors.
So, the answer to your question is: probably never.
Healthcare article at Kuro5hin