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


Forgot your password?
Software Technology

New Photo Fraud Detection Software 124

An anonymous reader writes "CNet is reporting that Hany Farid, Professor of Computer Science and applied mathematics at Darthmouth College, has developed a new version of his Image Science Group's photo fraud software now in use by the FBI and large media organizations. The current software is written in Matlab, but the new version will be written in Java making it much more readily available to local police and smaller media organizations. From the article: 'I hope to have a beta out in the next six months,' Farid said. 'Right now, you need someone who is reasonably well-trained to use it.'"
This discussion has been archived. No new comments can be posted.

New Photo Fraud Detection Software

Comments Filter:
  • but... (Score:5, Funny)

    by smash ( 1351 ) on Sunday February 05, 2006 @05:26AM (#14645077) Homepage Journal
    How do we know the fraud detection software isn't fraudulent? :D


  • by jibjibjib ( 889679 ) on Sunday February 05, 2006 @05:31AM (#14645089) Journal
    Can we adapt it to detect Slashdot article fraud?
  • open source? (Score:5, Interesting)

    by YrWrstNtmr ( 564987 ) on Sunday February 05, 2006 @05:35AM (#14645096)
    "Will you be able to get a copy of the Java-based version of the Image Science Group's applications? Probably not. One of the dilemmas of this type of software is that the more widespread the distribution, the more chance forgers will exploit it to their advantage."

    followed by -

    "...the software will be made freely available under an open-source license.
    "Taxpayers," he said, "are paying me to do this research and it needs to go back out." "

    Which is it?

    • Re:open source? (Score:5, Interesting)

      by aussie_a ( 778472 ) on Sunday February 05, 2006 @06:16AM (#14645172) Journal
      Those aren't necessarily mutually exclusive. He said access will be limited, most likely to police, FBI and other government agencies. But if those agencies want to enhance it, or make it better, they'll be able to do so under the open source license. The license may have a clause that limits who can get access to the software, but for those who are able to access it, typical OS rights are given to them.

      Or he might not know what he's talking about, and/or wanted to use the term "open source" for good free publicity.
      • The license may have a clause that limits who can get access to the software

        An open-source license can not have such a clause.
        • Re:open source? (Score:5, Insightful)

          by NathanBFH ( 558218 ) on Sunday February 05, 2006 @06:35AM (#14645207)
          Negative. The GPL doesn't have such a clause, but it is easy enough to modify the GPL (or any other license) to allow distribution of the source with the binary, and then restrict distribution of both. Granted, you may be using a different definition of the word 'open' than the software owner is. This scheme is more open than a closed-source solution where I sell you a binary with rights only to use the software and absolutely no rights to the source. While it may offend your OSS senses, it is, I think, a valid use of the term open source.

          If you really want to take the term 'open source' to the extreme, I could argue that even the GPL fails to meet some level of openness. The GPL restricts use of its source code on several accounts.
          • the gpl restricts re-distribution of gpl'd code, if it isnt redistrubuted under the terms of the gpl. i dont believe it restricts use, unless redistribution is defined as 'use'.
          • Re:open source? (Score:1, Informative)

            by Anonymous Coward
            Open Source is not some open ended concept, it has a clear definition: []
            • "When I use a word [...] it means what I choose it to mean. Neither more nor less. The question is who is to be master?" Humpty Dumpty (in "Through the Looking Glass").

              In real use, words (and phrases) mean what the speaker (writer) wants them to mean. They also mean what the listener (reader) hears them as meaning. (This is rather less the case in the context of formal documents of whatever sort - which is where your cited definition appears.) Definitions in dictionaries reflect that use (yup, I te

          • Well, the only truly open "license" is the public domain. Otherwise, the only real point of having a license is to delimit the use of the copyrighted work. And that's okay, one simply has to accept that the term "open" is not an absolute.
          • If you really want to take the term 'open source' to the extreme,

            I use the OSD (, nothing more, nothing less. And you can't restrict distribution. You can twist the meaning of 'opensource', I won't.
        • No, a Free Software licence may not have such a claim. Free and open source software are not always the same thing!
        • An Open Source Licenced piece of software can go to a community who will not distribute it outside their peer group, however. If no binaries make it outside that community, no source needs to be distributed outside that community, either. It cannot be done by edict as part of the license. But that's the only restriction.

          • Exactly. And that's what I mean. You can not put restriction but you can distribute the software only to a small number of persons. However these persons can do the same.
        • Gosh, a dumb elf has marked me as a troll.
          Darnassus will burn!
          BTW I've eaten him.
      • When you distribute source, but don't allot redistribution of source, that's closed source. I believe AT&T used to do exactly the same thing, back in the day.

        This seems like the professor is trying to limit access to his source in order to either keep the number of people with access to the softare low. This is probably for a couple of reasons; people could pick holes is his source and find ways to beat his code, or he just doesn't want anyone outside of government to have the program.

        He's throw
        • He allows redistribution of the source, merely to a limited amount of people (based on profession rather then numbers). I guess it depends which part of open source you think signifies something to be called open source. Altering and redistribution to particular people, or redistribution to anyone?

          Funnily enough, all open source licenses have limitations. So a limitation that says "only the following american government agencies" doesn't necessarily exlcude it from being open source, IMO.
          • Maybe, but this whole discussion is merely philosophical, since the source would be readily available for download very soon in P2P networks.
      • But if its government funded and tied to the GPL, i would think they would be bound to the requirement to give us a copy if we go thru the proper channels. ( like most other government funded software.. and all GPL code. )
        • by Anonymous Coward
          Who said anything about the GPL? Professor Farid certainly didn't.
        • No one is required to give you a copy of GPL software. The only requirement is that if someone _does_ give you the software, they must also be prepared to give you the source code.
          If they didn't give you the software in the first place, you're sol.
          In real terms, if the photo guy distributes his GPL'd software only to people who are required by their employers not to redistribute it, then chances are he is acting in the spirit of the licence.
          I wouldn't know how federal regulations might interact with this.
      • Re:open source? (Score:4, Interesting)

        by bill_mcgonigle ( 4333 ) * on Sunday February 05, 2006 @11:16PM (#14648074) Homepage Journal
        Or he might not know what he's talking about, and/or wanted to use the term "open source" for good free publicity.

        Odds are something got lost in translation. I met the guy a few years back and he's quite sharp and very nice and unpretentious. He gave me a copy of the paper this work is based on. I thought at the time he should commercialize it. Open source would be even better.

        Anyway, the paper was published and an algorithm should be able to be implemented by anybody with the appropriate skills. So, somebody could do a GPL version even if he doesn't.

        The company I was with at the time wasn't smart enough to accept his offer to collaborate on some research. Just as well for him, I say.
    • I'll be watching for the program in alt.binaries.warez.* hierarchy....
  • by icydog ( 923695 ) on Sunday February 05, 2006 @05:50AM (#14645126) Homepage

    As pointed out earlier, apparently the source code won't be released but it is open-source. Interesting.

    Anyways, also FTFA:

    Still, around 1 percent of accepted articles contain manipulated images that do significantly affect the results, said executive editor Mike Rossner. Those papers get rejected.

    So do they get accepted or rejected?

  • by Tsar ( 536185 ) on Sunday February 05, 2006 @05:53AM (#14645128) Homepage Journal
    Note that the submitter identifies the product as "photo fraud software" not "photo fraud detection software." This is quite apt, since the application will obviously cut both ways. Someone cooking a photo could simply run each version through this software, making minor tweaks until their "improvements" pass its inspection. If the software works the way it appears to, it would be the image manipulator's equivalent of a spell checker.
    • Ah, yes, a spell checker. Now there's something we *can* use to detect slashdot article fraud.
    • Or inverting the algorithm to "correct" a picture, it may not be that easy, but it does provide a guideline to those that want to make stuff up a way to validte it as real.
    • It may be possible to invent a type of apparel that would give a false-positive to the photo manipulation algorighm, so when you wore it, it would always detect a manipulation in the photo, even if they weren't actually doctored.

      They wouldn't be able to say 100% that it was an original, unretouched photo. That may give enough wiggle room to say that the glove didn't fit...

      I'm picturing a shirt/mask/helmet with a pattern on it that resembled jpeg blocking artifacts.
  • Now the CIA can outsource 1/2 their work... :P If we hack it into a video fraud detection package we can effectively get rid of the CIA all together...
  • In capitalist west scientist detects your airbrush fraud.

    In Soviet Union you airbrush scientist out of photo. Soviet_Union []

    The fun a government can now have with this package will be great.
    False positives to kill a story?
    Could real digital "abusing prisoners" images now be spun as a hoax?
    Just a few well placed reports as to the authenticity of any new digital images could kill a story?
    Or lure a leaker out to 'prove' the reality of the images only to

  • Fallibility (Score:5, Insightful)

    by Mikey-San ( 582838 ) on Sunday February 05, 2006 @06:44AM (#14645220) Homepage Journal
    From the article:

    "Right now, you need someone who is reasonably well-trained to use it."

    I would like to hope that if this software is going to be used for anything of consequence, that someone reasonably well-trained will always be using it. A system is only as good as its operator, ultimately.
    • It really depends on how hard it is to use. If it's the kind of thing where you feed it the file, and it spits out a percentage of how likely it is that the image is fraudulent, then perhaps you could have a regular cop working the system. Or at least a Cop who deals with computers on a day to day basis. I don't think you'd really have to know exactly how it works to get it to work right. I'm just worried that the error rate might be a little off.
  • by Jivha ( 842251 ) on Sunday February 05, 2006 @06:56AM (#14645232)
    As other commenters have already pointed out, the confusion over its open-source nature(is it or is it not?) is critical. Without the source code/algorithm being open-source and freely accessible to the public how can one trust its "judgement"? In a legal situation, an accused will always question the accuracy of the algorithm and the software.

    On a different angle, I wonder how soon before such detection capabilities will be available to consumers either as an installable plugin or web-based feature. Imagine being able to verify the authenticity of any picture on the web, ranging from that nude shot of your ex- to that impossibly perfect low-light picture taken by your photography class buddy ;-)
    • by jimicus ( 737525 ) on Sunday February 05, 2006 @11:03AM (#14645727)
      What you're saying is a rehash of the argument that "Information security isn't secure when the algorithm's security depends on its secrecy".

      Thing is, does the same thing hold true when you're talking about detecting fakes (say), as opposed to building strong encryption? If I announce "Well, we can tell this photo isn't genuine because this part which shouldn't be in focus is", I've effectively announced to any potential fraudsters who might be listening "OK, folks, you need to learn to get your focusing correct".

      Realistically, the only way such an algorithm remains secure is if it cannot be beaten even with a full understanding of how it works - and I would ask if such an algorithm even exists yet. If the algorithm is anything less than 100% effective, chances are it doesn't.
      • It it can analyze a 640x480 image and tell if it's a fake or not then it can be beaten.

        There are only so many pixels and so many combinations thereof that it'd quite simply have to be possible to make a fake image that meets all the criteria for a real one.
      • 'Information security isn't secure when the algorithm's security depends on its secrecy"'

        I think his point was that people will run images through this software and the software will say "fake" and users will believe it, even when it's wrong. The same problem is true whether closed or open source. They'll substitute the black box's definition of 'fake' for the real definition of fake because it's easier.

    • that impossibly perfect low-light picture taken by your photography class buddy

      Still object or moving object? Low-light scenes can pretty much be compensated by large aperture and/or long exposure. Large aperture increases depth of field, and long exposure blurs moving objects. He could also use a high speed film that is more sensitive to light but results in more grain in the image. With experience, you can find a good balance between those three factors and take perfect pictures.

      I'm sure you know all this

      • Large aperture increases depth of field
        Larger apertures (i.e. f/2.8) have a shorter DOF than small apertures (i.e. f/22).

        DOF info []
  • It determined that Pamela Anderson's boobs are FAKE!!
  • by Loconut1389 ( 455297 ) on Sunday February 05, 2006 @07:18AM (#14645259)
    Rewrite software written in powerful mathematical simulation software in java?
    What could possibly go wrong?

    And now, rather than processing an image in 30 minutes, it takes 30 hours, yay!
    • by titzandkunt ( 623280 ) on Sunday February 05, 2006 @08:08AM (#14645343)

      "...What could possibly go wrong?..."

      Well, memory leaks and array bounds probably won't go wrong ;-)

      Looking at some benchmarks [] for numerical processing using Java, it appears to stack up quite well agains C++ at least.

      Yeah I know, what exactly is being measured, are the benchmarks relevant, are any benchmarks relevant, blah blah blah. Just pointing out that the parent's postulated x60 slowdown is a trifle pessimistic.

      • exaggeration is what makes the point/makes it funny, no? =)
      • I saw a 60x slowdown in a journal paper describing this exact topic(Java vs. Fortran). Their summary was that it may be possible to get Java fast but by default Java is(was) slow.

        Look at "Java for Numerically Intensive Computing: from Flops to Gigaflops" or "Java for high-performance numerical computing". These both tell that better libraries(for multidimensional arrays) and relaxation of the floating point requirements of Java can speed up things a lot.
        • Yeah. If left to its devices, the JDK will do all numerical stuff in double-precision. This is absolutely diabolical for things like acumulating the results of a matrix multiplication, where intermediate result accuracy can be allowed to slide while the final total accuracy must be maintained. Who ever said that numerical calcs were easy? Java can be slow, but it doesn't have to be...
          • Those papers I listed describe the problems with java, including the need for precise exception handling which limits compiler code restructuring(like loop unrolling), among the commonly known things like bounds checking for all array accesses.
      • It looks like, at least for light direction computation, linear algebra is used intensively. If they use LAPACK for Java, then their program will at least attain the same efficiency as in MATLAB if not more efficient. MATLAB programs are interpreted, so they could not be more efficient than compiled code.
    • I hope that sand your head is buried in is nice and warm.
  • by Elitist_Phoenix ( 808424 ) on Sunday February 05, 2006 @07:42AM (#14645298)
    In related news NASA announced today that it would close its public picture archives ;)
    • The White House might have some explaining to do as various news agencies run every last picture on or from through the fake analyzer.

      The Administration has been caught numerous times photoshopping soldiers into and out of pictures. (And I'm sure Clinton did it too.) I hope a tool like this would encourage a little more honesty & a little less photochopping from the gov't PR people.

      /And if a program like this really takes off and is distributed to the media, I think it's fair to expect it

  • by pugugly ( 152978 ) on Sunday February 05, 2006 @07:58AM (#14645323)
    So - when can I expect to see this as a Firefox Extension?

    "Warning: This nude of Britinet Spears has been photoshopped"

  • ... when it's come full circle: Photo manipulation SW gets so good that it can fool the photo manipulation detection SW every time, so you're going to need human specialists to detect the kinds of manipulation that SW can't. Just hope the jury hasn't been replaced by a 12-member P2P system by then.
  • Apparently they consulted with's [] talented staff of professional photo retouchers while testing the software.
  • I wonder what would happen if the software was used to process photos from the original Apollo moon landing. I am of the belief that there is a great possibility that at least the original Apollo landing was faked. Subsequent missions I am not too sure about, but I believe at least the first one was a fraud. I wrote this page [] up many years ago when I was in 10th grade at school. After reading it again just now, however, I really think I need to update it for new facts I've discovered and general maturity of
    • The last pictures shadow looks how it should be -- The leg of the lander is angled, which causes the shadow to look off.
    • Your arguments sound awfully familiar to those already refuted []. In short: Stars aren't usually visible because of the fact that capturing those would need longer exposure, meaning the foreground would be overexposed (hint: Try taking a photograph anywhere with odd lighting conditions, then compare that to what you actually *see* - human eyes have pretty damn amazing dynamic range compared to cameras!) The odd shadows are mostly due to the fact that the surface isn't quite as flat as it seems, and the object

      • My main arguments were (and I point out again I wrote that when I was 15) regarding the photos, I didn't really put much stock in 3rd and 4th hand information such as the stars argument (I knew of the dynamic range rebuttal) and I didn't even bother mentioning the ridiculous "flapping flag" argument. There's one photo in particular that looks very, very faked. I'd tell you which it is but that'd require me opening a new browser and I can't be bothered :P I believe I highlighted it in my little rant.
    • Amateur HAM radio operators communicated with some of the Apollo missions. HAMS need to position their antennae correctly to do this, which confirms that the signal did come from at least the direction of the moon, and not from some California backlot.

  • $MATLAB/bin/mcc (Score:1, Informative)

    by Anonymous Coward
    If it's a matlab program, a C version is available now.
  • A matter of time (Score:2, Interesting)

    by BiDi ( 853932 )
    Just a matter of time before this application gets it's brother: "the simulator".

    It uses the same algorithms in a slightly different way: instead of checking for the signs of forgery it finds the tell-tale signs of modification and then reverse-modifies them to "what-should-be-there" to make an "original" modified image.

    The result will be an image that is ofcourse different only from mathematical standpoint - visual information will be the same. If that wouldn't be true I would love to have an application t
    • You are on crack. No mathematics is going to magically reconstruct pixel information that has been destroyed. If it does, I'm going to get straight to work on a program that starts with a blank hard drive and mathematically interpolates "what should be there", movies, software, whatever.
    • by pyite ( 140350 )

      It uses the same algorithms in a slightly different way: instead of checking for the signs of forgery it finds the tell-tale signs of modification and then reverse-modifies them to "what-should-be-there" to make an "original" modified image.

      What makes you think this is possible? Let's say I have a set of 15 numbers, {1, 2, ..., 15}. Now suppose I consider each of those numbers modulo 3. I now have a much smaller set of unique numbers, but multiple numbers in the original set map to the same new number. I ca
  • Photos have gradual changes in color (even when there are abrupt changes in the scene being photographed) and edits don't. But if people are trying to produce edited photos that need to be undetectable, it should be possible to write a filter that fixes edited photos. To do the equivalent of taking a photo of the edited photo. It's easy to seek for someone who isn't hiding. When they try to hide, it's a bit tougher.
    • This is quite a generalization. I think a skilled artist could fool this software quite easily. My inclination is to believe that it might compare things like lighting temperature, and overall color makeup between various components, but if you have someone that knows that they're doing, this isn't a big deal.
  • I wonder if the software could be modified to easily uncover the blacked out parts prevelant in many government documents. I suspect they would put up a big stink about this though.
  • by craXORjack ( 726120 ) on Sunday February 05, 2006 @12:35PM (#14646066)
    The light in the famous doctored photo that puts Sen. John Kerry next to actress Jane Fonda at a protest rally actually comes from two different directions.

    "The lighting is off by 40 degrees," Farid said. "We are insensitive to it, but computers detect it."

    Well even if that one is fake, at least we know that the one of John Kerry french-kissing Joseph Stalin is real.

  • I feel like an old curmudgeon here, shaking my fist and shouting "dupe! []".


  • I wonder if/when it will be used on the Oswald-with-gun photo that many (apparently starting with Oswald) claim was faked.

    The photo appears at the start of this wikipedia article on Lee Harvey Oswald.

    (Of course the article is the subject to disputes of its own. B-) )
  • by AgNO3 ( 878843 ) on Sunday February 05, 2006 @03:15PM (#14646684) Homepage
    Once you have the comp in done in your image editor with as little pixel and grain distortions as possible. Try the following. Note this does require a $5000 set of plugins [] Furnace (no not the After deffects set) and the $5000 program them run in. [] or [] [] (and yes a few others in the $100k+ range) In fact I might even do the whole comp in my film compositor with the use of some other tools. Anyway. Comp elements use historgram matching to match elements in the shot that should have the same color ranges. (You could do a color overlay in PS.) Ok now you have a good comp with good edges or edge blending and light wraps. You completely degrain the shot with furnace (each element seperatly degrained) You then regrain the shot as a whole. degraining and regraining work on all three color plates seperatly so when you regrain the shot it should be adding basically another blending of the of the colors making them uniform to what the original piece of film would have been. Now you take that file and do a film out of that. That is a very simlified break down of the technique. Many of the steps for each step are left out. My basic assumption is that the software looks for irregularities in the pixels deformations, areas of transition, and color offsets in the comp. I would love to go up against this software. (not that I would win but it would be fun to try as long as someone else is paying for the film outs and scans :-).) OH and if your source and destination or supposed to be film you probably want a drum scan not a CCD scan.
  • It appears that the program also detects software bloat -- it found unnecessary manipulations with some

  • Another example not mentioned is the OJ Simpson mugshot photos that were on the cover of Time and Newsweek at the same time, where they were the same photo but one adjusted the contrast/gamma so it was darker, more sinister. Not exactly a touchup/photoshopping but still a modification done on purpose to skew interpretations.

    I seem to recall some US guvmint propaganda as well where the picture showed a large group at a speech, denoting good turnout and agreement, yet when you looked closely you saw the same

You are in a maze of little twisting passages, all different.