Slashdot is powered by your submissions, so send in your scoop

 



Forgot your password?
typodupeerror
×
Crime Technology

Boston Police Chief: Facial Recognition Tech Didn't Help Find Bombing Suspects 235

A reader writes "ArsTechnica reports: 'While the whole country is relieved that this past week's Boston Marathon bombing ordeal and subsequent lockdown of the city is finally over, Boston Police Commissioner Edward Davis told the Washington Post that the department's facial recognition system "did not identify" the two bombing suspects. "The technology came up empty even though both Tsarnaevs' images exist in official databases: Dzhokhar had a Massachusetts driver's license; the brothers had legally immigrated; and Tamerlan had been the subject of some FBI investigation," the Post reported on Saturday. Facial recognition systems can have limited utility when a grainy, low-resolution image captured at a distance from a cellphone camera or surveillance video is compared with a known, high-quality image. Meanwhile, the FBI is expected to release a large-scale facial recognition apparatus "next year for members of the Western Identification Network, a consortium of police agencies in California and eight other Western states," according to the San Jose Mercury News. Still, video surveillance did prove extremely useful in pinpointing the suspects.'"
This discussion has been archived. No new comments can be posted.

Boston Police Chief: Facial Recognition Tech Didn't Help Find Bombing Suspects

Comments Filter:
  • by Anonymous Coward on Monday April 22, 2013 @07:49AM (#43514619)

    Rinse and repeat

    • by rudy_wayne ( 414635 ) on Monday April 22, 2013 @08:00AM (#43514679)

      If only all those millions of security cameras were as good as they are on TV. But they aren't. The images they produce are shitty and worthless. So they identified the suspects by having FBI agents sitting at a monitor and watching video over and over and over.

      But that won't stop the FBI from rolling out yet another billion dollar boondoggle facial recognition system.

      • by kannibal_klown ( 531544 ) on Monday April 22, 2013 @08:33AM (#43514863)

        What? You mean you can't "Zoom and enhance" a 640x480 video to the point that you can see the fingerprint left on a window 25' away

        Unfortunately, I know people that actually think that stuff is legit. Which of course leads to "fun" arguments / questions about "Why can't you do THIS, I see them do it on TV all the time."

        • by Psyborgue ( 699890 ) on Monday April 22, 2013 @12:18PM (#43516855) Journal
          There are actually some decent enlargement algorithms [weizmann.ac.il] out there that can do better than you would probably think possible. Take a look at the vision chart image. GIGO is an unbeatable law, but there's a lot more you can do with "bad" input than most geeks think. Dude below talking about license plates has the right idea. If the missing information is present elsewhere in the image, or if you have prior knowledge, you can use that to reconstruct portions of the image you want. The more repeating patterns in the image, the better.
      • by sycodon ( 149926 ) on Monday April 22, 2013 @08:40AM (#43514905)

        It should be noted that the old facial recognition software worked just fine. It was the fancy new computer based facial recognition software that failed.

      • by plover ( 150551 ) on Monday April 22, 2013 @09:37AM (#43515351) Homepage Journal

        Actually, even low resolution cameras can be really useful, under certain conditions. If the suspect stands still for a few frames, the images can sometimes be enhanced due to motion differences between the frames. The process is like anti-aliasing in reverse.

        In the video clips i saw on the news the suspects were walking, and the differences between frames looked too great to get the kind of data needed to interpolate.

        If you're interested in seeing this done in a non-fakey-CSI application, Thierry Legault is an astrophotographer who uses frame interpolation to produce amazingly clear shots of objects like the ISS. See his site here to learn more: http://legault.perso.sfr.fr/ [perso.sfr.fr]

        • This technique is commonly used in plate recognition software for example. Program identifies all frames with plate, crops/scales/aligns and then goes to town. You can have grainy 320x240 of a car where plate looks like a white line 3-4 pixel high and it will still be able to extract real number out of it.

      • So they identified the suspects by having FBI agents sitting at a monitor and watching video over and over and over.

        No, actually. They spent hundreds if not thousands of man-hours looking over the video, and then the guy with his legs blown off came out of sedation and wrote "I know who did it" on a pad, and the FBI was at his beside not longer after, getting a description of events and the suspect.

        The witness said one of them plonked the backpack down next to him and walked away. Looked him right in the

    • by kevingolding2001 ( 590321 ) on Monday April 22, 2013 @08:35AM (#43514881)
      Don't forget the 'uncrop'

      http://www.youtube.com/watch?v=KUFkb0d1kbU [youtube.com]

    • by Ksevio ( 865461 ) on Monday April 22, 2013 @09:03AM (#43515067) Homepage
      Did the software beep enough? It needs to beep for each magnitude you zoom in.

      Best if every face scanned is shown on screen next to the original and beep.
  • .. use Google Picasa :D
  • Computers are not magic.
    • Re: (Score:2, Insightful)

      by Anonymous Coward

      tell that to my users, please.

      And add that I am not a magician.

    • People who don't know any better have been known throughout history to explain just about anything as magic. Why should computers be any different? Now, if you disagree, I will "re-program" my computer to shoot lightning at you across the internet!

  • by bc90021 ( 43730 ) * <bc90021&bc90021,net> on Monday April 22, 2013 @08:03AM (#43514695) Homepage

    If I'm not mistaken, the CCTV footage was not as useful... what did help was the one man who took a picture of the bomber (unbeknownst to him at the time), and more importantly, the unfortunate man whose legs were blown off at the knees who valiantly gave an ID from his hospital bed.

    • by Xest ( 935314 ) on Monday April 22, 2013 @08:57AM (#43515017)

      Is there really something valiant and courageous about ID'ing the person who just turned you into an amputee?

      I'm not belittling the man, I feel awful for him as that's one of the most horrendous life changing things I can imagine happening, but I'm not entirely sure what heroic act this man has performed, he's done what anyone in his situation would do - the maximum he can to exact revenge.

      Perhaps this is a cultural thing, but the bravado being shown regarding people who did what anyone would expect them to do between this and the poor MIT officer who got shot dead without being given chance to defend himself strikes me as a little odd.

      I would argue, the heroes, if any, are those who rushed to the aid of the injured without knowing if they themselves could become victims of another bomb or attack as they did so, not the poor sods who died or are led in hospital beds - they're unfortunate victims. Is no one allowed to be a victim in America? Must every victim be made a hero whilst the real heroes go unnamed and unknown?

      Certainly I imagine that if this is what heroism is, the guy led missing his legs would rather be one of the unknown and unnamed.

      • Re: (Score:3, Interesting)

        by Anonymous Coward

        I imagine that someone waking up with their legs blown off might have more legitimately selfish things to worry about than catching the bomber such as whether they were going to live, how long they were going to be in hospital, or mourning all the favorite activities that would not be possible for a long time to come - if ever at all. I'd argue that for someone in that position to wake up and immediately want to talk to investigators is heroic. He may not have made a heroic choice to be involved in the

      • There's an unfortunate tendency to label "heroes" people who suffer some dramatic, tragic fate. The crews of Challenger and Columbia were hailed as heroes, when in reality they weren't any more heroic than any other space shuttle crew, just a lot more unfortunate.

    • by jellie ( 949898 )

      Wasn't it a combination of all of the above? The FBI collected video recordings and photos from all available sources, and identified two suspects. The FBI had one of the suspects putting the backpack on the ground right before one of the explosions, and also saw the two of them walk away from the scene afterward. That information was enough to pick those two and, for example, rule out the people identified by the NY Post and Reddit. But the images weren't clear enough, so they asked for the public's help f

  • Pinpointing the suspect's what?

    Oh wait, that's grammatically correct. It's a sad day when I start imagining /. grammar mistakes that aren't actually there.

  • by symes ( 835608 )

    The surveillance footage probably helped identify the bombers. But they were initially apprehended by an MIT campus security officer and while he may well have known who to look for it was his training and bravery that made the most significant impact in this case. So questions about the value of CCTV and other tech to one side, we mustn't forget there is a very important human element in amongst all. I kind of feel it is imporatnt to not lose sight of this.

    • Re: (Score:3, Informative)

      by paiute ( 550198 )
      The MIT policeman was apparently shot in the head while sitting in his car, not apprehending the suspects.
    • by dcw3 ( 649211 )

      But they were initially apprehended by an MIT campus security officer and while he may well have known who to look for it was his training and bravery that made the most significant impact in this case

      *Boggle* Did you make this shit up? If not, you need to disqualify the source as a reference for valid information.

  • by SpectreBlofeld ( 886224 ) on Monday April 22, 2013 @08:18AM (#43514783)

    Gotta wonder if it picked up matches for random people who are wanted for one thing or another, and if there will be follow-up investigations on those leads.

    And if so, if crowd-scanning will become a precedent...

    • Authorities have always wanted magically accurate identification. And yes, they want to use it everywhere. They don't appreciate the difficulties inherent in applying a system that can match people in high quality images against a database of a few hundred high quality images, to poor pictures against a database of over a million tiny, poor quality mug shots. Even if they upgrade all the photos, a system with a fantastically good 0.01% false positive rate will still find about 100 matches for every perso

  • WTF? (Score:4, Insightful)

    by Anonymous Coward on Monday April 22, 2013 @08:18AM (#43514785)

    "Boston Police Chief: Facial Recognition Tech Didn't Help Find Bombing Suspects"

    Thousands of paramilitary, guns, Humvees, helicopters, robots, hours and hours of lockdown of millions of people and the suspect went uncaught.

    A homeowner on a smoke break finds him.

    Who the fuck cares about facial recognition, I say arm the citizens and save money and time.

    • It works in Syria!

    • Re:WTF? (Score:5, Insightful)

      by rvw ( 755107 ) on Monday April 22, 2013 @09:36AM (#43515345)

      Who the fuck cares about facial recognition, I say arm the citizens and save money and time.

      Go to Somalia and find out how good that works out.

      • A gun is the difference between a citizen and a subject. I categorically reject any attempt to turn the former into the latter.
      • Arm and educate. Somolia lacks one of the two that the U.S. as a first world country really should not.

    • Who the fuck cares about facial recognition, I say arm the citizens and save money and time.

      Nice of you to promote your pet cause. Despite the protests of the NRA, you can buy firearms, even in Boston. Perhaps you're not satisfied with the citizen's choice to call the BPD instead of foolishly getting his head blown off. I might point out, that despite not playing Rambo, the actions of said citizen did lead to catching the perpetrators. I think that's a good thing - how about you?

  • Not really a fan of this technology - but my thought is this would be a good place to work on fine-tuning the system to increase the effectiveness. You have several RL image sources/raw footage and know what the result should be... time to work on debugging.

  • by swillden ( 191260 ) <shawn-ds@willden.org> on Monday April 22, 2013 @09:06AM (#43515087) Journal

    Current-generation face-recognition systems have a false positive rate of about 1 in 1,000 even when they have excellent images to work with -- high-resolution, well-lit, full-face frontal photos with no obscuring hats, glasses, etc. So even if CCTVs captured excellent images, if you're searching a database of tens of millions you're going to get a lot of matches. In a case like the Boston bombing it's okay if you get a few thousand hits because there is manpower available to sort through and narrow those down to the dozens which the (much more accurate) human eye/brain can't distinguish, and then there's manpower available to chase down each of those leads.

    When you reduce the image quality, though, make it grainy, at an angle, poorly lit, and throw in some baseball caps... forget it. You have to reduce the match threshold, and then instead of thousands of candidate matches, you have tens or hundreds of thousands. For that matter, consider the fact that humans can't deal well with those constraints, and we're social animals who devote a significant portion of our enormous brain capacity to exactly this task.

    • by booch ( 4157 )

      I'm not convinced that we'll ever have facial recognition software that will be able to identify anyone in the USA. The false positive rate would have to be below 1 in 300 million to be completely automated. That's a really high bar to achieve. In addition to your points about poor quality images and various angles, there's also the fact that people's faces change as they age. Other things can fool recognition software too -- facial hair changes, facial expressions, makeup.

      My guess is that something else wi

  • by 140Mandak262Jamuna ( 970587 ) on Monday April 22, 2013 @09:06AM (#43515095) Journal
    All I have used is picasa. And I have been impressed by its ability. It might have fizzled out in this instance, but this technology has real potential.

    I have loaded some 45000 pictures, almost all family pics, on to Picasa. Once I identify a face and tag it, it finds the same face in other photos. And as I mark yes/no for its findings, it improves remarkably. It is not confused by heavy make up worn by Bharatnatyam dancers. [google.com] It is finding the correct faces of 20 such dancers lined up facing the camera. It picks faces obscured in dark backgrounds, in out of focus pictures, faces occupying hardly 50 x 50 pixels. Faces at all orientations, including upside down. Half faces, faces with just one eye... It is really amazing.

    What is amazing is its mistakes. It mistakes mother for daughter and vice versa. Confuses brothers with sisters when they are toddlers but not when they are teens or adults.

    But this is forward match, going from a known face and looking for it in a crowd. Boston police is trying the reverse look up on a massive scale. It failed today. But like Lycos and webcrawler being upstaged when Google solved the reverse look up problem, some day the reverse look up problem will be solved. With parallel technology? Through GPU's running million forward searches simultaneously? But someday soon, the reverse look up will be solved and the automatic photo identification will work.

    • by gl4ss ( 559668 )

      it's probably more of a problem of there being too many hits, the grainy image doesn't have that much data and if you have to compare it to couple of million potential candidates.. finding it from a set of few hundred like you would with fb/picasa/whatever image tag is easy though. not because there's not enough processing power mind you, but because it's going to be hitting too many potential matches.

  • by sl4shd0rk ( 755837 ) on Monday April 22, 2013 @09:16AM (#43515177)

    So if the system doesn't actually catch bad guys, why do they still have it? Did they not save their sales receipt from spending all those tax dollars?

    • by gmuslera ( 3436 )
      The idea is to have high res cameras to be able to catch everyone, everywhere, guilty of something or not. Will be like checking up in foursquare every place you go. Or at least that is the idea they have. False positives will make life very interesting in the US.
      • there is an ongoing effort to intimidate the populace into mental submission to government and its police. This is a necessary step in the transition to a police state. Just as one of many examples, four decades ago, the phrase "lock down" was used for one place, a prison during riot or violence incident. Now it is used for schools and major cities.

      • The idea is to have high res cameras to be able to catch everyone, everywhere, guilty of something or not.

        In the same spirit as the USSR. They didn't have this fancy tech, but they tried. What many people don't know is that their law enforcement was terrible, in the sense of catching the bad guys. They simply wanted to sniff up everybody's butt. Some people like doing that, and will use anything to rationalize it. But honest-to-goodness law enforcement actually requires work, lots of it, and leads to fewer sweetheart contracts for useless equipment. American law enforcement is better is better precisely because

  • Don't get arrogant (Score:4, Insightful)

    by onyxruby ( 118189 ) <onyxruby@ c o m c a s t . net> on Monday April 22, 2013 @09:27AM (#43515267)

    What people don't understand is that for facial recognition software to work you have to have good quality cameras, images, a more static environment. This is why you hear about it being used in casinos is Las Vegas and elsewhere. In those environments you have high quality cameras with close range and good angles working against a smaller set of good pictures in a relatively static environment (people in casinos tend to congregate and not move around a lot). You also have staff with a distinct vested interest in watching out for their 'bad guys'.

    In a place like a large public venue you have lower quality cameras, far more people running around, worse angles and range and the environment is far more transient. The tool is being used in a completely different environment with far less support and far larger data sets to work with.

    It's like taking your Rav4 off-roading the Rubicon trail and coming way with the conclusion that off-roading is a bunch of hype. You've taken the tool (grocery getter) and put it to use for a job it was never meant for. Meanwhile your guy with the old Jeep knows for a fact that his tools works for the job because he uses it for that job on a routine basis, however he would be just as foolish to except his jeep to work as well as a daily grocery getter as a Rav4.

    Until the tools are put into environments that allow them to succeed, and with the hardware that they need they will continue to fail. You could call it a failing of the tool, however the tools and hardware are immature. Give it another five years and this would be a very different story. It's just technology advancing and there isn't a damn thing you can do about it short of getting hold of your politician and demanding reforms or limits on it's use.

  • Reddit was just a larger version of Maple Street.

    http://www.tv.com/shows/the-twilight-zone/the-monsters-are-due-on-maple-street-12606/

  • That's what they WANT you to think...

Think of it! With VLSI we can pack 100 ENIACs in 1 sq. cm.!

Working...