Hitachi Develops New Visual Search

Tech.Luver writes to tell us that Hitachi has developed a new visual search engine that can supposedly find similar images from within millions of video and picture data entries in around 1 second. "The technology assesses the similarity of images based on image characteristics presented as high-dimensional numeric information. The information is acquired by automatically detecting information regarding the images, such as color distribution and shapes."

Comments

Hmmm.... robotics?

[fyngyrz]

This is interesting to me - if it performs well - because this is one of the key missing elements for robotics; robots have a lot of trouble trying to match the environment around them to stored records of objects unless the environment is severely constrained. I'm not speaking of AI here (or at least, not yet) but just robots that would be able to clean your floor, carry your groceries, navigate in a burning building, walk your dog, tend your lawn. If they can classify images against stored images well, we're that much closer to generally useful and at least semi-autonomous robot devices.

Training might be a little annoying the first few times, but once you had a good database, you could replicate - or share via RF, that'd be freaky... neighbor's robot learns what a ferret looks like, now yours knows too - so that newer models were more and more informed right out of the box. Crate. Coffin. Whatever.

Add an associative database so that images normally found near other images which have just been found are searched first, and perhaps you could get the general search time down from the quoted 1 second, I'm thinking. One second is kind of pokey for a lot of robotic applications. But if the thing is in a kitchen, why would it need to be looking to recognize images that are found in a shipyard?

And I, for one, would welcome our semi-autonomous, environment recognizing, floor cleaning robot underlings.

Re:Hmmm.... robotics?

[dotpavan]

what I am curious about is the computing resources required to process.. other than the algorithm, is this one of the reasons which is delaying the emergence of search in the field of images/music/video on a commercial level? riya made some strides, but is still "learning"

Re:Hmmm.... robotics?

[lawpoop]

I'm not speaking of AI here (or at least, not yet) but just robots that would be able to clean your floor, carry your groceries... Well, you are talking about AI here. It turns out that it's relatively easy to make a computer that can beat humans in chess or do complex math equations, but something as simple as walking with 6, 4, or two legs, which a lot of really stupid organisms do, is really difficult. Something like distinguishing 'indoors' from 'outdoors' or a cloud bank from the bushes, seems way in the future.

My pet theory is that we don't have the right kind of device yet. A mind, the 'function' of an organic nervous system, is not a Turing machine. I don't really understand the math behind it, but Goedel's incompleteness theorem seems to show that a human mathematician can understand certain mathematical proofs that a Turing machine can never prove. Since all computers are a essentially a Turing machine, no matter how fast or parallelized they are, or how much memory they have, they will never be able to do what a human mind can do. So, maybe someday we will have artificial intelligence, or, a floor-washing robot, but we currently don't have the right kind of device that can do it.

Re:Hmmm.... robotics?

[Anonymous Coward]

The problem with this idea of storing images is that it could be pointless to store that much data. Your brain does not store images of say a tank. Your brain says "Does it have tracks?", "Does it have a turret?", "Does the turret rotate?". Well to figure out if it has tracks it would then search characteristics about tracks, then search for information about turrets etc. Once that is done you know it is a tank, now you need to know what kind. So you would need to know stuff about specific types of tanks like the Russian T-34 has a slanted front. Maybe search for a Russian flag, well then you have to figure out if the symbols on the tank are a Russian flag etc.

Most everything is made up for basic shapes. Some of those shapes may only be described by mathematics like for example, circles, triangles, squares etc. Now you just have to save how those shapes fit together to make a bigger shape but a shape in 3 dimensions.

Say for example you wanted to find a trash can. You would look for something that looks like a cylinder or a rectangle with a hole in the top. It would then have to be sitting on a flat surface. The top of the can may not be well defined due to the bag. Thus you have found a trash can. So rather then searching for images search for common shapes and combine those shapes to find what you are looking for. The problem is picking out the shapes.

Re:Hmmm.... robotics?

[fyngyrz]

Well, you are talking about AI here.

No. I'm not. There are robotic vacuums and lawnmowers right now. I'm just talking about giving them some eyes so they know not to mow your puppy or your child or your roses, or vacuum up your engagement ring. Teaching a robot firething not to step into a hole in the floor, and to rescue people before pets, and pets but not stuffed animals.

It turns out that it's relatively easy to make a computer that can beat humans in chess or do complex math equations, but something as simple as walking with 6, 4, or two legs, which a lot of really stupid organisms do, is really difficult.

Not so difficult that it hasn't been solved multiple times, multiple ways, including such variations as stair-climbing and running. Nothing to do with AI, either; just a progression from over-complicated attempts to solve using complex equations over the whole assembly to simpler approaches like fuzzy-logic based feedback systems that work right at the joints.

Something like distinguishing 'indoors' from 'outdoors' or a cloud bank from the bushes, seems way in the future.

Not if there's a good image-matching mechanism, it isn't. The concept of "looks like" is a very powerful one. That's what Hitachi says they've done here; we'll see if it lives up to the report.

My pet theory is that we don't have the right kind of device yet.

Keep that theory warm. Reality has a way of bringing the cold, fast and harsh. My pet theory is that a serial computer architecture can emulate anything, anywhere, given the proper code, enough storage and enough time to jump through all the hoops; to which I add, once you get it working, you can optimize the code and the hardware to do the job better until it is in the realm of the practical, if the investment is worth it. And for AI, IMHO, any investment is worth it. That's been the history of every solved problem so far, and I see no evidence that any solvable problem will be any different. And intelligence is solvable; after all, nature solved it may ways.

Re:Hmmm.... robotics?

[dj_tla]

All of the robotics problems you described - cleaning your floor, carrying your groceries, navigation, etc. - are AI problems. In fact, even the "(1) high-speed visual similarity search using two-step search clustering technology" that Hitachi is promoting could be considered AI. You see, in computer science, if a problem has some kind of real world application (even in theory) it goes under the moniker of AI, and this gets papers published. In my opinion, AI's just a rebranding of the rest of computer science; object orientation becomes "frames," complicated parsing problems become "natural language processing," robotic navigation is "pathfinding." But I digress!

Anyway, more to the point, Hitachi's technology doesn't do any learning, and thus, doesn't require training. It's also not particularly new technology, they abstract out some features for the images in their database, then search for images with similar features. This is probably oversimplification, I'm not a computer vision guy, but I'm relatively confident that Hitachi has not furthered the state of the art in computer vision. The reason I guess it's newsworthy is the second part: "(2) faster reading through optimized data allocation on an HDD." This isn't a new technology either, though. I'm not sure if TFA is really that newsworthy.

Robots have a lot of trouble trying to match the environment around them to stored records of objects unless the environment is severely constrained. This is mainly because robots are specialized. For navigation, infrared/laser/sound sensors are better suited, as they tell the robot how far away obstacles are. Robots that are concerned with identifying objects usually do not need to navigate.

Re:Hmmm.... robotics?

[kebes]

Your implication is that the human mind cannot be reduced to a Turing machine. I am in the other camp--who believe that the mind is subject to rigorous physical law, and that physical law can be expressed arithmetically (in principle), and so the human mind is a Turing machine.

Godel's theorem says that a consistent arithmetic system will contain unprovable truths. Put otherwise, such a system cannot be both consistent and complete. Thus the Godel counterargument to Strong AI (that human minds and computers are not fundamentally different) is that humans (e.g. mathematicians) can prove things like Godel's theorem, so we are able to "rise above" the arithmetic and exist in states of full proof and full consistency.

But I think there is a flaw in that logic (note: I am not a mathematician). The theorem doesn't preclude that a given arithmetic system (e.g. human mind) will be able to prove a truth that a weaker system ignored. Thus our ability to see certain truths doesn't mean that there are not other truths that are unprovable to us.

More fundamentally, no one has actually shown that the human mind is either consistent or complete (proving both would be required to show that we are not subject to Godel's theorem). The human mind is a computational device evolved to solve real-world problems, like escaping predators, rather than contrived ones, like mathematical proofs. It is thus in fact likely to be an inconsistent (internally contradictory) computational system. The human mind may be incomplete and inconsistent.

I agree that "true AI" will require vastly more computer power, and much more sophisticated algorithms than we have today. But the emerging evidence, from what I've seen, is that "true AI" can be achieved, at least in principle, by a Turing machine.

Re:Hmmm.... robotics?

[kebes]

The field of evolutionary psychology is attempting to do just that: to deduce what algorithms are working in the human brain. One of the end-goals of such research is of course to be able to generate artificial versions of those algorithms. If you're interested in such things, Steven Pinker's "How the Mind Works" is a fantastic and accessible read. It describes how things like vision are optimized for environments we evolved from, and tend to fail when put in contrived situations (like optical illusions). It also tackles the "adaptive advantage" of having emotions, and so forth.

Of course, it's easier said than done to actually transplant a biological algorithm into a computer. Even when you figure out the basic strategy ("it seems to be a neural net that responds to edges in a visual image and passes shape information along to the next module"), it turns out that the details are difficult (millions of years of evolution have adjusted the exact "weighting factors" to very specific values!).

Re:Hmmm.... robotics?

[lawpoop]

Describing the algorithms qualitatively is different than defining what they actually are. Evolutionary psychologists really don't do anything essentially different than say something like "Human beings are able to perceive a difference between indoors and outdoors, and clean and dirty. This allows them to keep their house clean, which provides a selective advantage in evolutionary terms, over and above an ape that can't tell the difference between clean and dirty, or indoors and outdoors."

They aren't really *defining* the algorithm ( and I don't think it's safe to assume that it *is* an algorithm as we know them, given the spectacular failure of AI so far ) , just describing it. Saying that "Humans can and do perceive 'indoors and outdoors' is different than describing the actual mechanism of how they do it. The actual functioning of the algorithm remains a black box. Which means then that you don't have any plans or blueprints for building a robot that can wash floors.

Re:Hmmm.... robotics?

[fyngyrz]

The whole idea of things being impossible based on hierarchies of understanding and/or proof is specious in the extreme. It is a dead-end philosophical backwater. Problems can be, and often are, solved without full understanding. Nature does this all the time; evolutionary algorithms can do it too. So it is irrelevant as to if we can understand AI, or not. The only relevant question is whether we can arrive at it in any way possible, and that question will only remain open until, and if, someone gets it done.

It is also useful to recognize that it is often true that there are multiple ways to solve the same problem. For instance, if you want to perform division, there's long division, which is clever and solves the problem relatively quickly, but you can also just subtract the divisor from the dividend and count how many times that can be done until the result goes negative. Both completely solve the problem and get you the same result. With regard to AI, it may be that we find a solution that is not the solution nature found for us, and it may be that it is trivially easy to understand. Or not. My point is that the legions of nay-sayers start with a lot of presumptions that have not been established as fact and go on to make these assertions on very shaky ground indeed.

I'm perfectly ready to say that we don't understand ourselves, and agree that we are intelligent. But that in no way leads to the presumption that we can't create intelligence some other way, or that we can't understand how it is done. Or that it might not be a more effective intelligence than that which we sport.

If nature can solve the problem - and it obviously has - then there are ways to solve the problem. If nature can do it with locally accessible materials - and it obviously has - then it can be done with locally accessible materials. What is lacking at such a point is merely technology. I fully expect full-on AI to be developed, and I see no known correlation that implies we'll have a good understanding of ourselves at that point in time.

I agree that "true AI" will require vastly more computer power, and much more sophisticated algorithms than we have today.

I think I can show you that this isn't so. If you agree, as you seem to, that AI can be embodied in an algorithm running in a Von Neuman architecture, then a slow computer should be able to solve the precise problems a fast computer can, it will simply hand you the result(s) later than the faster machine. Would you not agree that if the problem requires intelligence to solve, that the speed at which exactly the same, and entirely correct, answer is delivered is not a valid metric one could use to say intelligent or not? After all, one could (speaking generally) simply speed up the system (more memory, faster clock) and still get the same answer, perhaps now in the same amount of time; it's still not any smarter, just more convenient. And convenience in the sense of speed is a natural progression of technology.

From here, we can observe that there is no limitation in today's technology that says we can't put X amount of memory on a custom machine made with readily available tech, both ram and HD; additionally, any CPU can emulate any other CPU. So I say that there is no technological limit we face today that would stop Ai from functioning. Might be slow; but we can provide the hardware resources without question. And if it can be done slowly, hand it to the hardware folks and they'll optimize the hardware when they see what it spends most of its time doing, and it'll get faster. And faster, and faster... :-)

Conversely, I would expect that once the algorithmic issues are addressed, that we'll see intelligence - real intelligence - coming back to what many thought was incapable hardware. You might get your answer in eighty hours instead of a second, but if you get your answer... there you have it.

Re:Hmmm.... robotics?

[lawpoop]

Your implication is that the human mind cannot be reduced to a Turing machine. I am in the other camp--who believe that the mind is subject to rigorous physical law, and that physical law can be expressed arithmetically (in principle), and so the human mind is a Turing machine.

I'm not saying that the mind is not subject to physical law, or is not based on math. All I'm saying is that the mind is not a Turing machine ( though it probably would have to have a Turing machine in it somewhere ). It's a different *kind* of machine, not a super-powerful Turing machine.

Goedel basically showed that a Turing machine cannot do *all* the kinds of math that a human mind can do ( though it can do some). Not that a Turing machine lacks a certain amount of power, but just that it never will. It's just quantitavely the wrong tool for the job. It doesn't matter how much power you give it; the 'weakest' Turing machine is essentially the same as the 'strongest' one; it just simply can't do certain things. If a human is able to perceive and understand this, to know something that a Turing machine can't know, then the mind cannot *solely* be a Turing machine. This does not mean that the mind is not a different *kind* of machine, based on physical law, instead of some mystic hocus-pocus; it's just that it's not a Turing machine. My claim is that the mind is a qualitatively different kind of machine, not a Turing machine.

Goedel's theorem says that a consistent arithmetic system will contain unprovable truths. Put otherwise, such a system cannot be both consistent and complete. Thus the Goedel counterargument to Strong AI (that human minds and computers are not fundamentally different) is that humans (e.g. mathematicians) can prove things like Godel's theorem, so we are able to "rise above" the arithmetic and exist in states of full proof and full consistency.

But I think there is a flaw in that logic (note: I am not a mathematician). The theorem doesn't preclude that a given arithmetic system (e.g. human mind) will be able to prove a truth that a weaker system ignored. Thus our ability to see certain truths doesn't mean that there are not other truths that are unprovable to us.

I don't think the implication of Goedel's theorem shows that we 'rise above' the Turing machine, but rather that we have a qualitatively different awareness or knowledge that a Turing machine doesn't have.

Goedel's theorem is recursive. Any human mathematician can see that no matter how powerful the symbolic system is, the Turing machine will never be complete; there will be truths that the system can't prove. No matter how much you expand a particular system to show any truth a weaker system missed, there will be more truths that the newer, more powerful system missed. This process can go on ad naseum into infinity. A human mind can perceive this foray into eternity, but the Turing machine has no way of proving it. How could a human mind perceive something that a Turing machine couldn't, unless we had some component that was fundamentally different than a Turing machine?

What we seem to have that the Turing machine doesn't is meta-knowledge. We can see that any attempt to create a complete and consistent arithmetic system on a Turing machine will just lead to an endless series of more powerful systems that produce ever more elusive truths, and the process never ends. In this sense the Turing machine is 'myopic' -- it will never stop and say "Hey, I'm not getting anywhere with this; this is an infinite loop. No matter how powerful the system is, there will always be more truths that it cannot express." It's unable to know what it can't know, so to speak. However, as humans, we can somehow see the 'big picture', that no matter how powerful a system you make, there will always be another level of truths out there.

More fundamentally, no one has actually shown that the human mind is either consistent or complete (proving both would be required to sh

Re:Hmmm.... robotics?

[fyngyrz]

And what kind of awareness does a Turing machine have?

It knows, and can examine, its complete internal state better than you do, for starters. As you add peripherals, it can obtain information on the world outside of its hardware. The sophistication of this awareness grows both in complexity and in abstraction as the algorithms that deal with the data become more complex themselves. There's absolutely no reason to presume there is anything magical about awareness. A thermometer is more aware of the temperature than you are most of the time, and better than you are at it all of the time. That's a result of design. You can expect the same from AI, should we manage to cobble some up.

The attribution of awareness or consciousness to any sort of physical machine, Turing or otherwise, is a giant leap of superstition that atheists, or rather naturalists, are largely forced to make.

We need make no leap whatsoever. Nature has produced human intelligence through a process of incremental improvements. Reproducing or modeling natural systems is something we've turned out to be very good at once we understand them; there is every reason to presume that this is just one more of the same types of challenges. The leap of superstition is entirely yours — it is that awareness and consciousness are not perfectly natural consequences of particular combinations of processes and structures. In order to reach that conclusion, you have to imagine a being whose existence cannot be substantiated by the facts at hand.

But it makes for some ugly thinking.

And what might that be? I see only beauty, intriguing challenges, mysteries to be plumbed, problems to be solved. What do you see that is so ugly, eh? And why? What are you afraid of?

Re:Hmmm.... robotics?

[aragszxki]

Gödel's theorems have nothing to do with representing the human mind in any form. They cannot be applied to the human mind for the purposes of answering the question of strong AI. Basically, the only thing that Gödel's theorems do is carry the Liar's Paradox ("This sentence is false.") to the level of basic arithmetic. There is no magical process that proves or disproves anything about the human mind. The confusion stems from the fact that the mathematical terms "incomplete" and "inconsistent" seem to imply so much more when quoted in a non-mathematical context.

For anyone who is interested in reading further, Gödel's Theorem: An Incomplete Guide to Its Use and Abuse contains a thorough discussion of the issue.

I would like to believe that we will achieve strong AI one day. However, referencing Gödel's incompleteness theorems just because they sound appropriate at first glance does not give any argument scientific credibility.

Re:Hmmm.... robotics?

[mikael]

To implement a visual search engine you need to be able to perform the following:

texture segmentation - splitting up a picture into segments of distinct objects. In a panoramic scene, you want to split the picture up into objects such as sky, ocean, waves, beach, boats, pier, wall, people, animals. As a psychological experiment, you can show someone a picture , point to a particular point and ask them what the first word that the associate with that point is. Then you will see how every scene becomes segmented by our own vision systems.

Basic image segmentation is implemented using edge detection by Fourier Transforms (FFT, IFFT, DFT). This is a very computation intensive stage that is typically implemented using DSP's, GPU's or even dedicated ASIC's. Data used by the FFT can be in any dimension 1D (audio/radar), 2D (images) and 3D (volume visualisation). But to match the resolution of a human eye, you would need a 100 Megapixel floating point framebuffer.

texture classification - having identified the silhouette of an object, now attempt to match the contents to a particular object. Simple ways include colour histograms and silhouette matching. More advanced methods attempt to simulate the first few layers of the human retina using Gabor filters, Ring filters and Wedge filters.
But just to model a single type of retinal cell requires one or more FFT operations for an entire image. And
there are at least twelve different types of such cells. For efficiency precalculated results of sample images are generated (these are referred to as feature vectors) and then compared against the results of any new image.
For a really technical explanation of how human vision works have a look at The organisation of the retina and visual system

texture retrieval - the actual design of the search engine to retrieve images through content rather than just keyword:

QBIC - Query By Image Content. IBM's image retrieval database system

All of this has to performed for a single image. For an entire movie requires the processing of hundreds of thousands of images.

Re:Hmmm.... robotics?

[fyngyrz]

But we also know that intelligence is not material in nature,

No, you don't "know" any such thing. Barring the possibility of projects so black we've never even heard of them, no one seems to know what intelligence is. Least of all you or I. Claiming you know its nature at this point in the development of science is absurd.

Not exactly a rigorous definition. It doesn't give any idea of what you think intelligence is. You say mice are intelligent and rocks are not. What about ants? Ant colonies? Plants? Fungi? Bacteria? Disorganized large assorted collections of organic molecules?

First of all, it isn't a definition of intelligence at all. It's just a set of conditions that take into account intelligence or the lack of it, and natural systems as opposed to manufactured systems. I will say that even if I've picked a wrong example, these conditions still stand, you just need a right example to replace my error.

Having said that, though, I don't know what intelligence is. I am under the impression that I can sometimes identify its presence by manifestation of actions, however, at least within the realm of nature. So I sometimes know when it is. Plus there are organic hints; all my experience points to it being an emergent property of at least a moderate degree of complexity of neural systems. So if an animal has a decent collection of nerves - we're back to mice and higher animals - then the hardware may be there, and they are worth watching for displays of tool using, emotional outbursts, nurture, intellectually moderated defense and aggression, sharing, mutual support, empathy, self-sacrifice, and so on. For instance, when one fish continually pushes another to the surface because the other is paralyzed and cannot swim up to its food, that gets my attention. When a cat wants my attention and comes to meow at me because it is out of food, again, I pay attention. When a cat uses a mirror to locate and clean crud off its coat, I pay attention. When apes can learn to use signboards and computer systems to communicate complex concepts, I pay attention. Combinations of these things, especially in large numbers, make me fairly confident that what I am witnessing is a manifestation of intelligence. What is intelligence itself? No idea. What does it cause? Now that I have some ideas about.

Ants are borderline; ant colonies less so. Plants are not. Fungi are not. Bacteria are not. Organic molecules, in large summary collections, may be, depending on various factors; after all, that's one description of a brain, depending on just how disorganized you are implying. My middle son is pretty damned disorganized. :-) Once you get too general, you fall into the "can groups of atoms be intelligent?" trap; of course they can, that's what our brains are.

I don't disagree that it's presumptuous to call the present research being done AI

People use the term AI as a pointer to research seeking various aspects of the goal of AI, and I have no problem with that. That in no way should be confused with the mistaken idea than anyone has made any public announcements of having actually created anything even remotely resembling the target. Again, I except the vague possibility of black projects we may not hear about for decades.

Re:Hmmm.... robotics?

[Ceriel Nosforit]

Gödel's incompleteness theorem is only true for non-trivial formal systems. A trivial formal system on the other hand can prove just about anything, very much like we humans do with emotions and totally whacko notions related to religion, political conviction, and such. If we decide that emotions are trivial formal systems, we humans escape the restrictions discussed here.

Re:Hmmm.... robotics?

[mikael]

Here's a couple of articles I read:

Facts about the brain
Rods and Cones

There are around 125 millions rods and 6 million cones in each eye, with the percentages of each color/wavelength (red = 64%, green=32%, blue=2%)

No Sense

The human eye has 100 million neurons per per eye of five types, but there are only around 1 million neurons per optic nerve (arranged in bundles of 1000).

pr0n

[tronicum]

great! new way to find even more porn.

Re:pr0n

[kebes]

You may be joking... but I think rather than submitting sketches, the user would submit samples of things that already match what they want. For pornography, if a search engine were able to find images similar to those already tagged as "I like this," that would be a really sought-after search engine!

More broadly, if a search engine were able to find similar pictures, then you could narrow down to the result you wanted by submitting images that are close to what you want. For instance you may have found a thumbnail of the image you're looking for, but can't find a full-size version. Or you have a few pictures of airplanes, but want a whole bunch more.

Another way the technology could work is to present you with a series of candidate images, and then you click on the one that is "closest" to what you want. It then performs the search again, showing new candidates, and you click on the closest match again. If the search engine keeps showing you things closely matched to your last few selections, then this iterative process would quickly home-in on images of exactly what you want. In this kind of mode, the search engine could be using visual similarity as well as keywords and tags to figure out what kind of images you're trying to find.

Lastly, the idea of sketching an image might also work--at least for simplistic images. For instance I've often thought that the symbol-picker applet should, instead of listing thousands and thousands of symbols (which font should I look in?), it should have a box where the user can clumsily draw the symbol, and then display close matches (if you draw a circle it would show the degree sign, the letter 'o', the number '0', etc.). If it works, image-similarity technology like this could be a way to find the desired symbol. (It might work for clipart, too.)

Re:pr0n

[Sciros]

Yah I was just joking around.

Really though, suppose you don't have any images of two chicks riding a wookiee in a gladiator outfit. And say you know there's one out there. Well, I'll tell you, Alex Ross has a much better chance of finding that image with his mad drawing skillz. Of course, once he completes his "query," he's made himself the image he was looking for. So I guess it's kind of pointless. I forget where I was going with this anyway.

Will Google copy or buy this technology?

[bepolite]

I would think this would be a big and useful upgrade for http://images.google.com/

similar to Video Google?

[gstone]

From the rather less than opaque description in the linked article, it seems that this works is a hierarchical extension to a system known as Video Google. This system detects two-dimensional features in every image of a video sequence. Then uses hierarchical clustering to group together "like" features together. The centres of these clusters are used as "visual words". Scenes from the original video can then be characterised by which of these visual words they contain.

Using these words, search engine style indices and techniques can be used to make searching -- by supplying an example image area which can have its words computed -- quite fast.

The key bottle neck here is the clustering stage: reducing the original input of typically hundreds of features per frame -- multiplied by 25 frames per second by minutes, or hours, of video -- to a much smaller set of clusters. It looks like the work in the linked article is using a modified clustering algorithm which does not require all of the data to be in memory at once.

The TRECVID project is a challenge style exercise where groups compete to provide the best search results for a given set of queries where the search material is hours of video.

Document images

[zymurgyboy]

Pictures are fun, but I wonder if it would be accurate enough to locate similar images of documents (and to what degree). It would be really cool (for me anyway) if it could look at, say, a million pdfs or tiffs that don't have embedded text and come back with everything similar/identical.

I frequently have to create large collections of images from all sorts of file types -- some text-based, some graphics -- that get housed in a collection of images for easy, standardized review. If there were something that could avoid the step of extracting text from them, or later OCRing them and still end up with a searchable image collection, well, that would be exceedingly cool. It would cut the initial time outlay I have to devote to virtually any given project I have to deal with by 25 to 50%.

Re:Not as useful as it sounds...

[Jrabbit05]

But it could be used to create algorithms to find quality pictures, good photographs without viewing all of them.

Re:Not as useful as it sounds...

[FleaPlus]

For example: I want to find more cat images. I feed it a picture of a white cat. I am more likely to be returned results of white dogs than, say, tabby or black cats.

It seems it would be straightforward to implement something analogous to Google Sets, where you could supply a few photographs of what you're interested in (say, several cat pictures of various colors, or several white-colored pets). It could then learn which of the features were relevant, and add weigh to those in its search.

Since when can an ancient indian tribe ...

[SengirV]

... long since forgotten, be responsible for such innovative technology?

Saw it done 10 years ago

[mattr]

Some time between 1992 and 1994 IIRC when I was working at the photo/press agency Pacific Press Service in Tokyo, I saw a demo of a system created IIRC by NEC which searched 90,000 photos in under one second, based on a color freehand drawing you would draw on the screen of the EWS unix workstation on which it ran. Basically if you drew a horizontal blue mass at the bottom of the screen you would get a lake, etc. In other words you could search by rough photographic composition. I am less impressed that after over 10 years Hitachi was able to do something along the same lines.