Slashdot Log In
Video Surveillance Identifies Threat Patterns
Posted by
CmdrTaco
on Sun Dec 09, 2007 12:31 PM
from the threat-is-coming-from-my-pants dept.
from the threat-is-coming-from-my-pants dept.
Ponca City, We Love You writes "When the 2008 Olympic Games kick off in Beijing next year, organizers will be using a sophisticated computer system to scan video images of city streets looking for everything from troublemakers to terrorists. The IBM system, called the Smart Surveillance System, uses analytic tools to index digital video recordings and then issue real-time alerts when certain patterns are detected. It can be used to warn security guards when someone has entered a secure area or keep track of cars coming in and out of a parking lot. The system can also search through old event data to find patterns that can be used to enable new security strategies and identify potential vulnerabilities. IBM is also developing a similar surveillance system for lower Manhattan, but has not yet begun deploying that project. "Physical security and IT security are starting to come together," says Julie Donahue, vice president of security and privacy services with IBM. "A lot of the guys I'm meeting on the IT side are just starting to get involved on the physical side.""
Related Stories
[+]
Olympic Committee Chooses XP Over Vista 283 comments
Vinit writes "The popularity of Windows XP is still making things difficult for Vista. Now Vista has again suffered a major setback, with Lenovo (Olympic 2008' official sponsor) installing XP on it's machines to run the Olympic Games' vital PC-related tasks. Vista will only be used in internet lounges set up for athletes to use during the games."
This discussion has been archived.
No new comments can be posted.
The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.
Full
Abbreviated
Hidden
Loading... please wait.
What we all need (Score:3, Insightful)
Pattern recognition to identify threats, before trouble occurs.
Soon come the day when, we can finally arrest people, before they realise that they're going to do something criminal.
Re:What we all need (Score:4, Funny)
Parent
Re:What we all need (Score:5, Insightful)
The police try to find patterns in activity already, but it is far less effective for a relatively small group of people to look for patterns than it is for a computer with many cameras.
This is exactly what is already happening but faster.
When you are in public, you are in public.
Parent
Re: (Score:2, Insightful)
Re: (Score:2)
Re: (Score:3, Insightful)
Yes, there are cases where profiling is illegal. But in and of itself, profiling is *not* illegal. At le
Re: (Score:3, Insightful)
Re: (Score:2)
Do you have a citation for this assertion? More generally, any information that supports your contention that computers are better than people at recognizing crime patterns in real-time. Even more generally, that computers are effective at this at all.
Re: (Score:2)
I suspect research comes under the guise " Pedestrian Detection". Add/substitute 'moving vehicle', 'target identification', associate 'Artificial Intelligence for Homeland Security', google for the profile of a 'Fei-Yue Wang' and probably identify an emergent pattern.
CC.
Re: (Score:2)
That notion can go fuck itself in public; the issue is not that black and white, and you know it. Maybe what people are proposing is, get this: we need to redefine what 'public' is. Why should my right to privacy only exist within the square footage of real estate that I can afford to rent or own? Why should my right to privacy be limited if my private acts don't cause you any harm?
Re:What we all need (Score:5, Insightful)
I think the problem is we are redefining what public is.
20 years ago, there was no expectation whatsoever that being in "public" meant your every move would be tracked by government officials potentially hundreds of miles away, and then stored for all time. That's not what "public" meant. People had an expectation that yes, anybody who was around you could potentially be watching you, but that kept it a relatively level playing field because you could pretty easily identify any threats to your privacy and avoid them if you like. If you were walking down an empty side street and needed to quickly adjust your belt because your pants were too loose, you could look around and do so without fear that cops are watching ready to jump you for "reaching for a concealed explosive" or even "intent to expose oneself in public" or whatever other nonsense law they can come up with.
That is the expectation we have always had for what "public" means - yes, you can be watched, but only by those around you, and that means that you can easily watch them back. Being able to be watched - and recorded - by someone many miles away is not what "public" means to me or anybody else. That's an intrusion, just like any other. You are being watched by people who are not there. And you have no idea what they're thinking or doing, even while they can watch your every move. It's a completely one-sided relationship where the other side has all the power. That's scary. And it's the exact opposite of what "being in public" is all about.
We don't need to redefine what public means, we need to take back its original meaning. Nobody should be allowed to watch a space that they do not own (ie. a public space) without being physically present.
Parent
Re: (Score:2)
Re: (Score:2)
They say that if you're doing nothing wrong, you've got nothing to fear, but everyone does something wrong (e.g. downloading music). With the surveillance nation, you could theoretically be caught for wrongdoings that you didn't even realize were w
Re: (Score:3, Insightful)
When being in public entails having your every move watched and recorded and profiled, that's more like being on private property, or a prison.
Beyond a certain point, making something bigger or faster or stronger in just one aspect pushes it over a line to where it becomes something very different - spin a propeller and it turns around, spin it fast enough and you suddenly have powered flight. The connection of
Re: (Score:2)
When I am in public, I do not expect that everybody is filming me and keep records of me whatever I do. I expect to have MORE privacy then why I am in a non-public place, like at work.
I am not paranoia anymore, I know that I am followed.
You OTOH are a bit TOO positive about all this and that is suspicuious, so we will ke
Who do they protect and serve? (Score:2)
Re: (Score:2)
Oh come off it! This isn't Minority Report or crimethink. It's a way for security to monitor high traffic, high risk areas. It issues warning and allows security to prioritize their time and respond better to interruptions. When you're in these places, you're in public. You're not in the privacy of your own home or anything like that. You're on public streets. By going out into public you've already given up
Re:What we all need (Score:4, Insightful)
Parent
Re: (Score:2)
Tracking Real Threats (Score:2)
But the big thing you say is that security is only being added in "high-risk" ar
Re:What we all need (Score:4, Insightful)
That makes perfect sense in the world of twenty years ago and the world we still mostly inhabit, but pervasive electronic surveillance threatens to change the meaning of statements like these. If you want to maintain the same rhetoric, make sure the words mean the same thing -- i.e., stop surveillance from de facto changing what it means to be in a public place versus a private place.
If you accept that "public place" means "a place where a detailed, permanent record of every action is captured and archived by the government," then you should rethink whether we want to have any "public places" at all. By that definition, perhaps only congressional chambers, courtrooms, jail cells, and the immediate vicinity of police officers should be public places.
Parent
Re: (Score:2)
Re: (Score:2)
As for me, when traveling through areas like Manhattan, I would much rather have a surveillance system like the one the article describes in place than an armed national guard member on every corner.
Given the choice, I would rather have the armed guard. First, they are visible and obvious; there is a gun behind the camera just the same. Second, they actually have a chance to *interact* with people and learn about where they are working. It has a chance of becoming less of an "us-vs-them" thing as I stare at this gal's tits from the safety of my video screen and wonder about all of these "perps" walking around, and more of a "how can we work together to make this safe" thing. In an area I lived outside
I suppose... (Score:2)
Re: (Score:2)
b) IBM sold a lot of video surveillance and face recognition hardware to China to monitor the "terrorists"
Anyone else see a pattern here?
ah, yes, /this/ stuff! (Score:3, Informative)
From what I understand, though, there's a nontrivial amount of hardware involved to process the video, and though that may be less of an issue these days with better computers, I'm wondering just how many CPUs they will be throwing at how many different video cameras for this.
And I'm sure it's imperfect and prone to false alarms and such, but that's why you put human beings behind it instead of machine guns, no?
Re: (Score:2)
Re: (Score:2)
Recognition logic (Score:3, Funny)
if (hoodie || foreign) police.respondto(camera.location);
Re: (Score:2)
if (Falun Gong || Human rights protestor) police.respondto(camera.location);
Good *old* IBM (Score:4, Insightful)
The important thing is, just like they had no idea their technology was helping make the holocaust more efficient and were just making a buck, it's completely unimaginable that the Chinese might continue to use it to crack down on dissidents afterwards.
so, are there any stats (Score:5, Interesting)
Re: (Score:3, Insightful)
Guys security is good. Raping the constitution, disregarding human rights, and doing a number of other unsavory things to attempt to get it isn't. However, something as common sense as this
Re: (Score:3, Informative)
Did the article make a point of saying this was an anti-terror tool specifically?
Yes.
http://www.nytimes.com/idg/IDG_002570DE00740E18002573A9007A49A5.html?ex=1354683600&en=e991061bf11b2f3e&ei=5088&partner=rssnyt&emc=rss/ [nytimes.com] a sophisticated computer system to scan video images of city streets looking for everything from troublemakers to terrorists.
Re: (Score:2)
Sounds a lot like this (Score:2)
This is such bullshit (Score:3, Insightful)
These systems have been tested before, particularly in England, where Thatcher's government paid a shitload of money that could have been used for something useful, and the only useful thing they got out of it was well-designed studies that demonstrated that these screening systems don't work.
Here in Manhattan, we had a video monitoring system set up in the labyrnthine Columbus Circle subway station for a couple of years. It also had no effect on crime. (Nor did it have any effect on the cops beating up innocent people, who happened to be black.) The City took money that could have paid for more police (hopefully honest ones) and spent it on video toys instead. Duh.
Now we're getting these digital cameras all over NYC -- even though we have good data from England, from our own pilot programs, from the Atlanta Olympics, and elsewhere, that they don't do what their promoters claim. What it demonstrates is that a huckster can sell hundreds of millions of dollars worth of useless digital junk to unscrupulous politicians accountable to a hysterical public and campaign contributors as long as it has blinking LEDs and they say the magic word "terrorism."
I challenge anyone to cite any scientific evidence, any pilot program -- not some security "expert"'s opinion -- that there are any computer "patterns" that can identify "troublemakers" or "terrorists".
Stop and think. The London suicide bombers walked on the subway with backpacks full of explosives. Innocent people go about their business on the subway all the time wearing backpacks. What pattern is there that a digital camera could spot?
The only good news in this story is that we Americans are finally ripping off the Chinese for a couple of hundred million dollars, which is good for the balance of trade. This is known in economics as the broken window fallacy.
Maybe we could sell them the Brooklyn Bridge too -- oh, wait, they already own it.
Re: (Score:2)
Suppose you drop the image recognition part and instead recognise people using personally identifiable information that is captured wirelessly from their mobile phones and RFID. That would give
Re: (Score:2)
At this time, we just don't know how effectively the resulting data could be searched for unusual behaviour. Nothing of this scale has been done yet.
That's my point.
But it could work, at least in principle, because all of the technology issues involved have already been solved for other problems.
I don't agree that it could work in principle, with any technology that we have now or in the forseeable future. What evidence do you have? I put out a call for evidence.
The one thing that otherwise-intelligent techies miss is that computer-level technology problems are easy (given a blank check). It's the other problems that are hard, such as: how do you tell whether somebody is a terrorist? How do you tell from watching him in a crowd?
The remaining issue is how well you can automatically distinguish between a terrorist and a regular person when all you know about each is everywhere they have been in the last few years.
Well, yeah. That's the problem. If you discover a
Re: (Score:2)
You have much more information about each person in a crowd than a CCTV picture. You also have data about everything else they have done. That's how the hypothetical surveillance system works: it doesn't detect terrorists, it d
The face of a terrorist (Score:2)
At this time, we just don't know how effectively the resulting data could be searched for unusual behaviour. Nothing of this scale has been done yet. But it could work, at least in principle, because all of the technology issues involved have already been solved for other problems. The remaining issue is how well you can automatically distinguish between a terrorist and a regular person when all you know about each is everywhere they have been in the last few years. Being flagged with a false positive could prove rather inconvenient.
This is the big problem. Terrorists are actually quite rare. There is therefore very little information to input on them and most of it is likely to be statistical anomaly. There was an AI test at one point getting a piece of software to recognize images with tanks in them. They had a relatively small training set but the software did really well with it. Hit it with some real data and it got essentially random results. Why? In the test data, the photos with tanks and without tanks were shot on separate da
IBM is not the only one (Score:2)
They are in fact the only supplier that has delivered fully digital IP-CCTV for Casinos in United States. Casinos tend to be quite picky when it comes to surveillance. IndigoVision also did the Olympics in Athens etc. I do not work for them, but I have lived in Scotland and are aware of their business.
The also technology similar to IBM for detecti
picks out all the people with "slanty eyes" (Score:2)
Nothing new here (Score:2)
The Evolution of Real ID? (Score:2)
This sounds kind of like the early stages of such a system, except that it doesn't immediately know who you are.
I have a suggestion for IBM ... (Score:2)
Maybe they could save some money on expensive computer hardware and use some of those picture-sorting dogs from the next story.
Good use for the Beijing Olympics (Score:2)
Re: (Score:3, Insightful)
If a human gets it wrong, with some luck hopefully his partner, or commander may get it right and make a better choice. No it d
Re: (Score:2)
Hate to burst your bubble, but companies are amoral. They have one sole priority: making money for their shareholders.
If that means selling high quality computers at good prices, they will do it. That also means, if theres a need from another country to do XYZ job for gobs of money, so be it.
It is our choice of customers to choose who or who not to associate with.
Re: (Score:2)