Visual Search Engine Tracks Stolen Images 223
Barence writes "A new visual search engine could help photographers keep track of their photographs whenever, and wherever, they appear on the internet. The TinEye search engine allows users to search by uploading a picture rather than typing in a keyword. It then conducts a pixel-by-pixel search across the internet, flagging all instances of that image even if it's been cropped, merged or digitally altered in some way. It's not just for copyright enforcement though; 'it's being used by researchers who need to find where an image came from to provide attribution, even people who are trying to find out who people are in old photos.' It's currently in beta, but you can try it out."
Kind of Misleading on the Old Photo Identification (Score:5, Informative)
"it's being used by researchers who need to find where an image came from to provide attribution, even people who are trying to find out who people are in old photos."
This may be nitpicking but I read the FAQ and it does not, in fact, claim to be able to accomplish this unless that exact same 'old photo' is posted elsewhere on the internet:
Can TinEye find alterations of a query image?
Yes. As long as they are alterations of the same query image, TinEye can find them and include them in your search results.
Note that search results are ordered by 'relevance' (i.e. how well the result images match your query image), so image alterations are typically found at the end of your search results.
How does TinEye work?
TinEye uses sophisticated pattern recognition algorithms to find your image on the web without the use of metadata or watermarks.
TinEye instantly analyzes your query image to create a compact digital signature or 'fingerprint' for it. TinEye searches for your image on the web by comparing its fingerprint to the fingerprint of every single other image in the TinEye search index.
So this example they list of the soldier must rely on the fact that the website contained the same exact image that the people had of the old soldier they were looking for. I can't expect it to take any image of Person A and return every single image (past & present) of that person. That's ridiculous.
I would expect that to work out very infrequently as I'm not aware of any huge digitized databases of old photos or even newspaper microfiche. Hell, I have postage stamp-sized photos of my grandparents with people who nobody knows who they are. I don't think this tool could help me.
Re:Kind of Misleading on the Old Photo Identificat (Score:5, Insightful)
"it's being used by researchers who need to find where an image came from to provide attribution, even people who are trying to find out who people are in old photos."
I think in this context, it's pretty obvious that the software's not trying to discover who people are, or who shot the photograph. It's the researchers who use this tool. If you have one website without attribution or other names, and you search for other pages, you might find a different page that has the same image along with more information.
Re: (Score:2, Informative)
it does work pretty well for example i searched for this:
http://imagecache2.allposters.com/images/pic/BRGPOD-WM/BRGWM-158277_72_48~The-Great-Wave-of-Kanagawa-from-the-Series-36-Views-of-Mt-Fuji-Fugaku-Sanjuokkei-Posters.jpg [allposters.com]
it did find the actual great wave, to be true after a ton of images that had replaced the poster with other posters from the same site but it did find them which was pretty good and would be useful for research
eg if you had a section of a photo and you wanted to find the rest etc.
Re: (Score:3, Interesting)
Now what would be handy would be if it could somehow sort them chronologically (maybe using the metadata, or maybe if the server will give the date-modified on the picture...). That would reduce the amount of searching if you knew you were going for the oldest known copy, e.g. you wanted to know where it originally came from + whatever info there was about the picture that might not be quoted elsewhere.
Re: (Score:2)
Re: (Score:3, Informative)
You have to login to read the FAQ. If anybody wants to avoid jumping through the hoops, here's the FAQ as a gif [imageshack.us]. Sorry about the resolution, you'll just have to pick a good zoom level...
I think everybody understands... (Score:5, Funny)
that the real purpose for this is to find the rest of sets ;)
Re: (Score:2)
Re: (Score:2, Funny)
Embedded Codes (Score:2, Interesting)
Unfortunately... (Score:2)
Re:Embedded Codes (Score:5, Informative)
"rather than a brute force pixel by pixel search"
They're blatently not pixel by pixel comparisons... look at the tech, don't listen to the woman! If it was pixel based then an image saved using two different implementations of jpeg wouldn't match up. It's probably more likely that a map of lines, shapes, patterns etc in the image is built up, and then they are what's compared. This means images that are different sizes, have different light/colouring (such as a high quality scan vs poor quality) and colour depths, but are of the same thing, can still yield results.
Err... or is that not what you meant by pixel by pixel search?
Re: (Score:2)
Re: (Score:3, Informative)
Well, they don't actually do a pixel-by-pixel search... they index them and create digital fingerprints, then it does a pixel-by-pixel on your "search query" image and compares its fingerprint with the fingerprints in the database. Pretty neat.
Re: (Score:2)
One small beef though: Most steganography/watermarking I've seen focuses on the artist using the software to add a hidden mark in the image, and then verifying that the hidden mark also exists in an alleged rip-off.
It says nothing about entity doing metadata cataloguing automatically extracting the marks and putting the decoded marks in an easy-to-query database. You know, making the data searchable. Even if that's data that is supposed to be hidden, you know, to ward against this "watermark tampering" thin
Re: (Score:2)
There is an image processing technicuq called the Feature vector [wikipedia.org].
This can be anything from a color histogram to a compressed FFT of the image. MPEG-7 files have texture descriptors [slashdot.org] built in.
A color histogram may not be of much use if someone alters the overall appearance of the image (color to monochrome or sepia tone). Silhouettes might not work if parts of the image are cropped or composited with another image. Monochrome texture segmentation and classification may be the only method that would work.
Funny thing, but I just shifted a bit a pixel. (Score:3, Insightful)
The least significant bit of each pixel. Oh, and now it appears that this tool doesn't work. (At least, I would suggest it isn't that good, I could be wrong. The article appears to suggest that it is that good, if you can take a photo on your phone of a painting, and then find an article on that painting...)
Oh well, I guess people still haven't learnt that the old ways of copyright are only hanging on through inertia.
Oh, and queue the predictable (and correct) responses about how you can't "steal" digital images. To steal a photo or a picture, you would have to take a physical copy belonging to someone, and deprive someone else of that physical copy, without their permission. (And the word "steal" doesn't appear to appear in the article, added to provoke page views I guess.)
Re: (Score:2)
Oh well, I guess people still haven't learnt that the old ways of copyright are only hanging on through inertia.
This.
Re: (Score:2)
Oh, and queue the predictable (and correct) responses about how you can't "steal" digital images.
Pedantically, you are correct, but you that's about as far as it goes.
You can pass off someone else's work as your own. You can deprive them of income, and make income off images that you do not have permission to use. Is that better phrasing?
Re: (Score:2)
People who violate copyright don't understand the "theft of labor" part, but it doesn't really matter as we know they are hiding behind excuse that simply because they are the greed ones who don't want to pay someone else for their work.
Oh, sure... we'll get the "try before you buy" types complaining, but how does that apply to a photograph?
Re: (Score:2)
Theft of labour? No, I can't say I do understand that... you still have your labour, and you still have the product of it. I just have a copy.
I might have gotten it for "free", but it didn't cost you anything, either. You're exactly the same after I copy it as you were before: you did x amount of work and you have a copyable product as a result.
Re: (Score:2)
Not if the product of your labor is only valuable when it is consumed during a certain window of time, or put to use in a certain way. Many such labors are like airline seats. Once the opportunity to sell it passes, it's gone. When someone else runs off with your work and distributes it outside of any arrangement with you, they are eroding the value of the work. Your take on things betrays a very disconnected, uninformed understanding of the way that a photograph can carry
Re: (Score:2)
I'm not talking about irreplaceable commodities (airline seat). I'm talking about easily copied digital media. Way to mis-apply my argument though.
Re: (Score:2)
No, YOU'RE missing the point. Twice, now. The labor that a photographer puts into certain sorts of work, and the market for it, IS irreplaceable. It passes with time. When you rip off that work and distribute it, you are stealing the opportunity that the photographer has created by performing that labor. The photographer, for example, spends a lot of money and
Re: (Score:2)
Well said.
So I can go at 4:00 in the morning to a specific spot that I found after days of looking for the best place to take a photo and set up my camera (which cost thousands of dollars for the best results, which was an investment into a business selling art that some people might find enjoyable) so that I get the perfect sunrise shot, and someone wants to copy it because, hey, I still have the original, so it's all good!
Re: (Score:2)
Yes. And your "fans" who don't care to pay for your work are helping to convince you not to in
Re: (Score:2)
While I can "build" many books from the original one in matter of seconds, it takes a lot of time for you to build many chairs just like the first one, and it's your problem, not mine.
Re: (Score:2)
I don't normally respond to ACs, but I think there is a genuine misunderstanding about this.
Yes, there is a difference, but it's a moot point - both works required your time and labor to create, the book obviously takes a LOT longer to create, yet a book, a paperback, sells for a tiny fraction of the cost of a well crafted,
Re: (Score:3, Insightful)
Yes, there is a difference, but it's a moot point - both works required your time and labor to create, the book obviously takes a LOT longer to create, yet a book, a paperback, sells for a tiny fraction of the cost of a well crafted, hand-made chair.
You focus on the one completely irrelevant part of the discussion: the time spent creating the book vs. the chair. It's completely irrelevant because "book" is not essential to this discussion. You can replace "book" with any other copyrighted creation, including one that took about 1 minute to "create".
Copyright only makes fabulously wealthy (by and large) those artists who have created compelling material... enough so that many people are willing to purchase, at a fraction of the cost it took to make the original product, the work of the artist. They're buying labor and talent and skill, which is the same thing you're buying when you buy a hand-made chair (I'd hope, anyway, as hand-made furniture tends to be quite expensive).
Not really, not any more. These days advertising and exposure has a lot more to do with it than how "compelling" the material is.
No, you're paying for the labor, skill, and talent... and people who put more effort, skill, and talent into their creations are rightfully awarded with more than those who put little effort into their work.
No, you're paying for the advertising. For an example, there's lots of bands o
Re: (Score:2)
In the digital age, that argument just doesn't hold water any more. The writing of the novel now can easily be the building of it as well. Whether or not you get a hard copy published more or less depends on how large you want your distribution to be.
So, what's your argument going to be once we have true 3D printing available in every house? No, it's not going to happen in the next 5 years, but it's coming, eventually. So what happens then when someone is able to buy a chair, have it scanned into the printe
Re:Funny thing, but I just shifted a bit a pixel. (Score:5, Insightful)
The least significant bit of each pixel. Oh, and now it appears that this tool doesn't work.
Yeah, how about you just watch the video on their website before suggesting that what they do would be as retarded as comparing the values of each pixel. It's surely closer to cross correlation, meaning it's nothing like comparing pixel values but more like correlating the image's space-frequency components.
By the way, does anyone have any clue what information they store and compare? They obviously don't cross correlate your search image with every image in their index every time you search, so what could they possibly store that would allow them to correlate images?
Re: (Score:2)
It's surely closer to cross correlation, meaning it's nothing like comparing pixel values but more like correlating the image's space-frequency components.
It almost certainly just IS cross-correlation, with a little algorithm to find the correlation maximum somewhere near the middle of the picture. This, of course, is simply another way of comparing pixel values. It's probably much, much faster, because so many established FFT algorithms exist for very fast multiplication-and-add operations, but it's essentially just pixel comparison.
Re: (Score:2)
Re: (Score:2)
They probably don't avoid it. On the order of 10^5 cross-correlations can be done on medium-res pictures per second with a fast machine, especially one with dedicated video hardware. They might drastically narrow their search space by having a library of perhaps 100 "basis set" or "canonical" pictures (faces, cityscapes, mountainscapes, ocean photos, animals, etc.) with varying colors and features which stand out, and do the first set of correlations with those. Depending on which of the basis set the targe
Re: (Score:2)
Re:Funny thing, but I just shifted a bit a pixel. (Score:5, Informative)
Oh, and queue the predictable (and incorrect) responses about how you can't "steal" digital images. To steal a photo or a picture, you would have to take a physical copy belonging to someone, and deprive someone else of that physical copy, without their permission according to SlashDot, but not the English dictionary.
Pet Peeve of mine: That's not the definition of "steal". It's only the SlashDot conventional wisdom. It's really not that hard to look up words on the internet. Here's a link to a dictionary [m-w.com].
Steal:
1 a: to take or appropriate without right or leave and with intent to keep or make use of wrongfully
Appropriate:
3 : to take or make use of without authority or right
Re: (Score:2)
The point is, to "steal" something you have to take it, or appropriate it (i.e. make use of it). The original owner no longer has it (or has use of it). Copying isn't stealing.
Re: (Score:2)
The linked page doesn't say anything about 'steal/stolen'... neither does the lil intro video clip on the linked page. Looks like slashdot headline that talks about that. Equivalent would be to describe the Google search engine as a tool for tracking people who have stolen text from your website... it's hardly an all round view of what the thing does, is meant for, or is mostly used for (which is obviously porn).
Re: (Score:2)
All the better — Slashdot's mainstream is finally realizing, that, indeed, copying may be equivalent to stealing, even if no tangible property changes hands.
Re: (Score:2)
No! There's a million arguments that completely miss the point that prove otherwise! Doesn't matter if you're "depriving on a permanent bases" the creators rights to control copy/distribution of the creation, as long as they get to keep the original copy of the creation, nothing is removed... *cough*
Anyway, what have the slashdot masses got to do with the actual article?
Re: (Score:2)
According to the FAQ, it's still able to identify pictures that have been colour adjusted, cropped, and even sometimes if it has been slightly rotated or had text added (or missing).
This impressed me... (Score:5, Interesting)
I found quite a different result. I nabbed an old photoshopped pic I did a few years ago, and uploaded it. TinEye came back with two results, being the two source images from the photos. That's impressed the hell out of me.
Gatesfeld search results [danamania.com]
For the full size photoshopped version, Gatesfeld [danamania.com] if you want to try the search yourselves.
Re: (Score:2, Funny)
Directly under his post I guess, assuming you're using a threaded or chronological view of posts.
(But yes, you beat me to it)
Interesting for the big boys... (Score:2)
Interesting for the big boys, but not so much for the amateur or even professional freelance photographer.
What are you going to do if someone ripped your pics from Flickr and claims them? Exactly -- not much.
Re: (Score:2)
Send them a cease and desist letter, and then sue them, just like the big boys?
Presumably if your someone who actually cares enough to check and see if someone is copying your photo, your also someone who cared enough to have it documented when / where you took it, or at the very least that you actually took it. Or as applied to art... you have the original copy.
seems like its not a difficult case to win...
Re: (Score:2)
I spend a lot of time cruising Flickr for good looking landscapes that I can use for slideshow backgrounds. When you spend a lot of time looking for the same style of image it doesn't take long to find people rippin
Re: (Score:2)
Yeah I hadn't thought of it that way before, but you're right, f**k it, lets shut the internet down and stop developing any technologies that can't be used to stop people ripping your pics from flickr. You're so right that there's no one else in the world who's interested in developing image recognition algorithms for anything other than protecting their snaps on flickr.
gugh
Re: (Score:2)
Re: (Score:2, Insightful)
Interesting for the big boys, but not so much for the amateur or even professional freelance photographer.
How does this NOT help the small photographers? It's exactly those guys who don't have the resources to find people using their content.
What are you going to do if someone ripped your pics from Flickr and claims them? Exactly -- not much.
Are you saying then when someone steals your image you have no recourse available? With this site you can find who's using it. What you do about it is up to you. And content owners do have recourse. They can contact whoever's using their content and let them know they're in trouble, then offer to work out a solution.
For example, a small food service business asks a pri
Re: (Score:2)
Yeah and get a whole lot of melons :-)
What would be really cool (Score:5, Funny)
What would be really cool is if you could upload a transparent 1x1 pixel image and it returns every image on the internet
Yes
Re: (Score:3, Interesting)
What would be even cooler is if you could search for transparent 1x1 pixels.
Re: (Score:2)
When I was a kid, we didn't have image search engines! We had to make our OWN transparent 1x1 pixel images! Kids these days...
Re: (Score:2)
When I was a kid, we didn't even have pixels, transparent or otherwise. We had to "draw" transparent dots on little pieces of paper with pen knives.
Re: transparent 1x1 pixels FTW! (Score:2)
You mean my unpublished Perma-Alpha site that uses these for spacing could get slashdotted? Hooray!
Re: (Score:2)
Re: (Score:2)
Given the number of pages that were designed using the 1 x 1 transparent gif workaround, it might end up returning the entire Internet itself.
Logo hunting (Score:5, Informative)
So to test it out I grabbed a couple of logos (AIG, Slashdot, Bluesquare, Nike swoosh) and found that what it will do is find scaled down images or ones of lower quality but it won't handle significant colour shifts. So AIG for instance have a blue logo but sponsor Manchester United where their logo is displayed on a red background, the Nike swoosh I tested had a white background and all I got was basic black on white swoosh elements.
Now with photos this is less of an issue as major colour shifts are unusual but it does mean that for commercial and design art its not really as applicable.
Re: (Score:2)
Hmm, that's actually pretty interesting, I wonder if they could make a colour-insensitive search that worked off the outlines? It would serve a significantly different purpose but still be useful, I think.
For that matter, I wonder if it would totally miss if someone ripped off a colour image but grayscaled it because it looked neater that way...
Re: (Score:2)
Now with photos this is less of an issue as major colour shifts are unusual
Of course, if you wanted to hide from a tool such as this, you could probably do so by shifting the colors a miniscule amount (say, on an RGB image, adding one point of red, green, and blue), which in most cases would hardly alter the image, but likely make it drop off the radar.
I'd imagine that you could do the same thing by altering even just a few pixels this way as well.
Comics (Score:2)
What a great idea (Score:4, Funny)
Re: (Score:2)
I see where you got your nick :)
Good Start (Score:2)
This is a good start and definally has the obvious applications. Hopefully, if this is successful, i.e. people use it, work on more complex systems can be created.
It would be really neat to find pictures with a certain symbol on them or even my face.
well (Score:2)
does this detect hidden images in images as well?
Terms of Service: (Score:3, Funny)
By using Idée's TinEye website you signify your agreement to the following terms and conditions, which may be updated by us from time to time without notice to you.
Submission of pornographic or illegal files is strictly prohibited. Do not submit any file that can be construed as pornography or is in violation of any law.
No porn searches?
Failure to comply with these terms may result in termination of your TinEye account at any time, without prior notice and at Idée's sole discretion.
Ahhh.... okay.... don't search for porn, or we might not let you search for more porn from that particular account. Gotchya. Hehe.
-
Re: (Score:2)
out of curiosity:
Why do you want to search for Pr0n you already have?
Re: (Score:2, Informative)
http://tech.slashdot.org/comments.pl?sid=648951&cid=24643811 [slashdot.org]
Re: (Score:2)
Re: (Score:2)
I dare you to upload a picture of a peach.
What is pornography? (Score:4, Interesting)
How the hell am I supposed to know what their company considers pornography? Can I search for The Joy of Life by Henri Matisse? [artquotes.net]
The company is based in Toronto rather than some ultra-conservative U.S. state; that gives me an epsilon more confidence the company won't take the "nudity = pornography" stance. Still, I wouldn't be surprised if a search equivalent to a risqué ad campaign in Europe would get you banned.
Re: (Score:3, Interesting)
Re: (Score:2)
Can I search for The Joy of Life by Henri Matisse?
Moot point, you already know who to credit ;)
Re: (Score:2)
The Joy of Life was simply the first thing that came to mind. (I am not an art connoisseur.) Thank you though for your much better example.
Search by Text, 2D, 3D, or even Doodles! (Score:2, Interesting)
I'm missing the file date (Score:2)
If you're going to track where a photo came from, I'd expect a timestamp to be useful.
So is this a crime against freedom? (Score:2)
This brings to mind an interesting question; guys like RMS, and even our own IdontBelieveInImaginaryProperty here at slashdot like to rail about how copyright is basically a scam, a crime against the public because it restricts "sharing".
Photos are copyrighted too. If this tech were for tracking copyrighted MP3's, there'd be howls of indignation here. It'll be interesting to see the reaction on the copyright side of the argument. If we don't see the standard "this is anti-freedom" arguments, it'll be intere
Re: (Score:2)
The difference, in my mind, is simple. I have taken plenty of good pictures in the past (a few are here [imageshack.us]). If somebody wants to rip off one of my pictures for their background, I'm totally ok with it... in fact I'm flattered (please do check them out and tell me if you like 'em ... thanks :). If they submitted one to a photo contest claiming it as their own, I'd be rightfully upset. Translating this into the world of MP3s, if somebody wants to download a song and listen to it that should be ok too; it isn't
Re: (Score:2)
I'm a photographer. I have lots of images on the net. If someone wants to use one as their chat icon (that's happened) or desktop background, I take it as a compliment.
If someone takes one and claims it as their own and/or tries to sell it, then there's a problem.
I'm all in favour of going after people who try to sell pirated music or movies.
Bye-bye pseudonyms (Score:2, Interesting)
And hello false-positive and derivative lawsuits (Score:2)
i can't wait till they implement a DMCA notice-bot and start shutting down every single *chan, every single image mashup community, deviantart, etc. with false positives, and the identification of those patterns within larger images.
The "free speech for sale" cover art comes to mind.
Re: (Score:2)
Someday, Anonymous Cowards may be easily identified by the style and content of their posts. Many trolls who think they are safe now will be outed using methods similar to this picture searching.
robots.txt (Score:3, Interesting)
My robots.txt excludes access to my huge collection of images.
So, either one can prevent discovery by this tool in a very simple way, or it ignores robots.txt. Which is it?
Registered users only? (Score:2)
Maybe I'm still not completely awake, but I can't seem to see anything other than a login form. Do we need to be a registered user to use this search engine? Or is it because of its beta status?
Perfect 10 et.al (Score:2)
What All /.'ers Want To Know (Score:2)
It will never find my images! [ref] (Score:2)
Tm
Re:Small letters (Score:5, Informative)
I suggest that if you're not going read TFA and can't be bothered to do any further searching of the site you are slamming that you read this page:
http://tineye.com/terms [tineye.com]
which specifically states the opposite of what you claim might be there.
The letters aren't that small.
Re: (Score:2)
Re: (Score:2)
better off if trying to escape it slicing images up into tiny squares and rearranging them in a table grid or something... it seems to be able to recognise slight differences in the image well enough.
Re: (Score:2)
Assuming human "readability" isn't a concern to you, why not just RAR it, password it, and put it on a filehost instead? If human "readability" is important, you obviously won't be scrambling the picture to begin with.
Re: (Score:2)
Who said anything about scrabling it??? Have you never built a jigsaw? Never seen how you can put pieces of an image next to each other so you can see the whole thing???
Re: (Score:2)
even automagically - just add some random minor alteration to pixels in statistically homogeneous sections of your foto
Easy, except for, you know, how it doesn't rely on exact matches, but finds images with minor alterations too.
Re: (Score:2)
What about cropping?
Re: (Score:2)
Re: (Score:2)
They don't do a pixel-by-pixel "search", that would be dumb and apparently people are misunderstanding here. They're using some kind of fingerprint, which sounds pretty similar to your histogram comparison. Also, FWIW, a histogram is generated by doing a pixel-by-pixel analysis of the image.
Out of curiosity, I'd be interested in hearing what your histogram comparison does if the image were colour-reduced (say you set the least significant bit of every channel of every pixel to 0, for example). That would ma
Re: (Score:2)
It's a good component of a larger recognition system, but many images that are VERY different will have similar histograms.
Re: (Score:2)
That's one case of the idea being worth nothing, the implementation being everything. Sometimes it's the other way around, it depends.
Can't wait till Google buys these guys though.
Million dollar ideas (Score:3, Interesting)
Steps 1 and 2 of your plan are n
Re: (Score:2)
This system is very impressive because they have managed to really increase the scale that they can create the vocab.
That's partly because this isn't trying to identify shapes and objects in the picture. It's just fingerprinting the entire image and indexing that.