New Photo Fraud Detection Software

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.'"

Comments

but...

[smash]

How do we know the fraud detection software isn't fraudulent? :D

smash.

Very good idea, but...

[jibjibjib]

Can we adapt it to detect Slashdot article fraud?

open source?

[YrWrstNtmr]

"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?

[aussie_a]

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.

Re:open source?

[NathanBFH]

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.

Re:open source?

[bill_mcgonigle]

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.

This article likes to contradict itself

[icydog]

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?

Same shot from a different angle...

[Tsar]

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.

Re:Same shot from a different angle...

[jibjibjib]

Ah, yes, a spell checker. Now there's something we *can* use to detect slashdot article fraud.

Fallibility

[Mikey-San]

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's a fraud, because my blackbox software says so

[Jivha]

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 ;-)

what could go wrong?

[Loconut1389]

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!

Re:what could go wrong?

[titzandkunt]

"...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.

T&K

In related news...

[Elitist_Phoenix]

In related news NASA announced today that it would close its public picture archives ;)

A Java Version

[pugugly]

So - when can I expect to see this as a Firefox Extension?

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

Pug

Re:It's a fraud, because my blackbox software says

[jimicus]

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.

MMMM. My first test to beat this would be.

[AgNO3]

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 http://www.thefoundry.co.uk/ Furnace (no not the After deffects set) and the $5000 program them run in. http://www.eyeonline.com/ or http://www.apple.com/shake http://www.d2software.com/ (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.