Slashdot Mirror


DARPA Developing Video Parser

coondoggie writes with an article in Networkworld about a disconcerting DARPA project. From the article: "If a picture is worth a thousand words, the scientists at the Defense Advanced Research Projects Agency would like to make that about a billion with a new software intelligent program. DARPA this month said it will detail a new system it would like to see built known as the Visual Media Reasoning (PDF) program. The main idea is to develop an advanced software program that can 'turn 'dumb' unstructured, ad hoc photos and video into true visual intelligence.'"

29 comments

  1. Use biological *informed* systems. by BWJones · · Score: 4, Interesting

    Why not design a connectomics informed system that mimics the neural retina and visual system? Something that takes the results of research like this and uses true biologically informed computing to do what neural systems are good at and silicon based systems are not so good at? After all, what they are looking at is a system that works like a retina works (more like a video camera and not a still camera), so why not go to the biology which is really good at comparing like streams of information and making like or not like decisions.

    More traditional background on retinal design and research can be found here.

    --
    Visit Jonesblog and say hello.
    1. Re:Use biological *informed* systems. by doublebackslash · · Score: 2, Interesting

      Why not design a connectomics informed system that mimics the neural retina and visual system? Something that takes the results of research like this and uses true biologically informed computing to do what neural systems are good at and silicon based systems are not so good at? After all, what they are looking at is a system that works like a retina works (more like a video camera and not a still camera)

      The eyes do not "see" in the sense of processing information. They turn light into nerve impulses. Ho-Hum. We've got that, in fact this isn't about that at all. They are dealing with already captured data anyway.

      so why not go to the biology which is really good at comparing like streams of information and making like or not like decisions.

      Yeah, biology is really good at making decisions. Shame that not a single person on earth has a system that can replicate it. Nothing even close. Biology is fantastic, but it falls over flat when we try to replicate it on a computer. This is due in large part because a brain tends to have more "power" to throw at the problem by many orders of magnitude. It is also differently abled compared to a computer, we can't just scale up and hope to emulate it, need the special hardware-software combo that is our beloved wetware.

      Let us say that a bio-technical solution was somehow brought to bear on this problem, someone wasted a wish perhaps. Then you get... what? What biological process or technique can be used here? The ability of mamals to discern individual objects in a scene? Then how to classify them? After a scene has been broken down by the biological side how is it turned into useful information? The mind would have to hand off some information at some point. Images suck because computers can't do that well yet (hence this project) and tokens representing actions and objects seems impossible for quite some time, it would resemble the complexity of a human mind.

      Nothing in biology can be applied to this problem directly, only perhaps simple ideas applied rigorously. Stop spouting your favorite rubbish.

      --
      md5sum /boot/vmlinuz
      d41d8cd98f00b204e9800998ecf8427e /boot/vmlinuz
    2. Re:Use biological *informed* systems. by BWJones · · Score: 3, Interesting

      The eyes do not "see" in the sense of processing information. They turn light into nerve impulses. Ho-Hum. We've got that, in fact this isn't about that at all. They are dealing with already captured data anyway.

      Actually they do process information. The neural retina is like a miniature parallel supercomputer at the back of your eye that does initial signal processing from the photoreceptors through the over 50 kinds (200 in other invertebrates) of neurons.

      Nothing in biology can be applied to this problem directly, only perhaps simple ideas applied rigorously. Stop spouting your favorite rubbish.

      No offense friend, but I can't figure out if this is a troll or that you are simply uninformed here. Biological neural systems are *very* good at discriminating differences in data streams. Nested neural systems then further refine those abstractions and you get more advanced logic. The problem in the past has been discerning what those connectivities are as most current models of neural connectivity grossly underpredict the biological reality of neural circuit complexity.

      --
      Visit Jonesblog and say hello.
    3. Re:Use biological *informed* systems. by doublebackslash · · Score: 1

      When it gets down to it I don't think that we have the processing power or knowledge to emulate the biological processes and even if we could we would then have to understand their output.

      --
      md5sum /boot/vmlinuz
      d41d8cd98f00b204e9800998ecf8427e /boot/vmlinuz
    4. Re:Use biological *informed* systems. by BWJones · · Score: 1

      We certainly have not in the even very recent past, because we did not understand how biological systems were actually constructed/wired. This project is an attempt to solve that problem and one of the potential outcomes is a general purpose knowledge of connectivity that can be applied to computational problems.

      --
      Visit Jonesblog and say hello.
    5. Re:Use biological *informed* systems. by doublebackslash · · Score: 2

      Perhaps I treated you unfairly then. I'm going to give your links a good read. Seemed, at first rub, like more of the same old. Now not so much.

      I don't mind being wrong, after all. Means I get to learn something new.

      --
      md5sum /boot/vmlinuz
      d41d8cd98f00b204e9800998ecf8427e /boot/vmlinuz
    6. Re:Use biological *informed* systems. by RKBA · · Score: 1

      The retina certainly does handle a small amount of processing and classification that could probably be mimic'ed with some analog logic, but the bulk of image processing is done in the much more complex visual cortex(although you do mention "visual system" in passing). Also remember that the human brain is an immensely complex analog processor. Analog logic can be simulated with digital logic of course, but digital is much slower depending on the level of precision required. Neurons have a digital component when sending an impulse along their axon, but the timing of the digital pulses depend upon biochemical and electrical analog activity (ie; a neuron is an integrator among other things).

    7. Re:Use biological *informed* systems. by Anonymous Coward · · Score: 1

      Yeah, biology is really good at making decisions. Shame that not a single person on earth has a system that can replicate it.

      Give me one month with Jewel Staite, and I guarantee I'll replicate biology...

    8. Re:Use biological *informed* systems. by Douglas+Goodall · · Score: 2

      The easier way to apply biology to this is to hire all the unemployed people in america to process pictures and categorize them And do whatever analysis is required. Since the human brain is so damn powerful and hard to replicate,, why not use some of them? Just think, if you employed 100,000 people as analysts, what a powerful decision engine that would be, and probably cost less to develop and run than some kind of blue-sky solution that DARPA might pay big bucks to some company to develop.

    9. Re:Use biological *informed* systems. by Anonymous Coward · · Score: 0

      Yes, but the retina isn't extracting meaning. It's processing data, not information.

    10. Re:Use biological *informed* systems. by Anonymous Coward · · Score: 0

      Let us say that a bio-technical solution was somehow brought to bear on this problem, someone wasted a wish perhaps. Then you get... what? What biological process or technique can be used here? The ability of mamals to discern individual objects in a scene? Then how to classify them? After a scene has been broken down by the biological side how is it turned into useful information? The mind would have to hand off some information at some point. Images suck because computers can't do that well yet (hence this project) and tokens representing actions and objects seems impossible for quite some time, it would resemble the complexity of a human mind.

      http://www.foxnews.com/story/0,2933,331672,00.html

      I bet the army knows more than the industry.

  2. Parse this! by ColdWetDog · · Score: 2

    Turn dumb, unstructured, ad hoc photos into video intelligence.

    Cool, if it works, it could compress YouTube into about 60 seconds.

    --
    Faster! Faster! Faster would be better!
  3. Will it ... by Anomalyst · · Score: 1

    recognize Mohammed, when it sees him.

    --
    There is no right to feel safe thru security vaudeville at the expense of everyone's freedom, privacy and tax money.
    1. Re:Will it ... by Anonymous Coward · · Score: 0

      It will be programmed to recognize all undesirables (colored people).

    2. Re:Will it ... by White+Flame · · Score: 1

      I'm colored white! Oh no!

    3. Re:Will it ... by Anonymous Coward · · Score: 0

      You might pass a quick hash comparison, but then a full string comparison must be performed to confirm undesirability, so you might be okay.

  4. More irony by Paul+Fernhout · · Score: 2

    Contrast: ""VMR will be an enhanced capability to generate the intelligence required for successful counter-insurgency and counter-terrorism operations" the agency said."

    With: http://www.pdfernhout.net/recognizing-irony-is-a-key-to-transcending-militarism.html
    "Likewise, even United States three-letter agencies like the NSA and the CIA, as well as their foreign counterparts, are becoming ironic institutions in many ways. Despite probably having more computing power per square foot than any other place in the world, they seem not to have thought much about the implications of all that computer power and organized information to transform the world into a place of abundance for all. ... There is a fundamental mismatch between 21st century reality and 20th century security thinking. Those "security" agencies are using those tools of abundance, cooperation, and sharing mainly from a mindset of scarcity, competition, and secrecy. Given the power of 21st century technology as an amplifier (including as weapons of mass destruction), a scarcity-based approach to using such technology ultimately is just making us all insecure. Such powerful technologies of abundance, designed, organized, and used from a mindset of scarcity could well ironically doom us all whether through military robots, nukes, plagues, propaganda, or whatever else... Or alternatively, as Bucky Fuller and others have suggested, we could use such technologies to build a world that is abundant and secure for all."

    That said, it seems like a cool project technically, with multiple uses in civilian applications.

    --
    A 21st century issue: the irony of technologies of abundance in the hands of those still thinking in terms of scarcity.
  5. This is a job for... by Anonymous Coward · · Score: 0
  6. Finally a topic where the goatse link is on topic by roguegramma · · Score: 1

    Finally a topic where the goatse link is on topic.

    --
    Hey don't blame me, IANAB
  7. Really? by mcguirez · · Score: 2

    "If a picture is worth a thousand words, the scientists at the Defense Advanced Research Projects Agency would like to make that about a billion with a new software intelligent program."

    I hope the software intelligent program better language parses.

    --
    When you hear hoofbeats, think horses, not zebras
  8. for the low price of one dollar! by Anonymous Coward · · Score: 0

    for one dollar i'll let you put your dick in my mud bucket

  9. Academic project stuck in the labs? by Anonymous Coward · · Score: 0

    Sounds like another academic project, hopefully not one destined to be stuck in the labs forever. Such is the nature and difficulty of computer vision research.

  10. Uwe Boll... by Anonymous Coward · · Score: 0

    ...might finally catch a break.

  11. Ambitious by supermariosd · · Score: 1

    Seems like a very ambitious project. The implications would be vast for intelligence if done correctly. I'm imagining a program parsing through terabytes of satellite data....

    Whatever advances come out of the project might also be applied to doing broad semantic studies of how people use video sharing platforms like YouTube. That would be good for better HCI and, of course, getting more relevant content. Exciting and scary stuff.

  12. Did I just flash? by toddbert · · Score: 1

    Maybe now I will get an Intersect in MY head!

    --
    "When half of your head is metal, having a few screws loose takes on a greater meaning". - Jack
  13. Re:Already done by tehcyder · · Score: 1

    The NSA already has an A.I. that can do this. Why is the taxpayer paying twice?

    The amount of financial fraud is mind blowing.

    And in other news, the NSA has discovered A.I.

    --
    To have a right to do a thing is not at all the same as to be right in doing it
  14. Intelligence Gathering, Circa 700BC ? by LifesABeach · · Score: 1

    Couldn't the members of the NSA do something that Sun-Tzu wouldn't yawn at?

    Couldn't the members of the NSA evaluate a working version of the Sorting Hat used commonly at Hogwarts? I know it sounds goofy, but why not take this concept to the engineers in the BACK OF THE ROOM, the ones that are muttering, "What would it take to make a Sorting Hat?"