Slashdot Mirror
DARPA Testing Numenta's Brain Tech
lousyd writes "CNN Money reports that DARPA and the National Geospatial-Intelligence Agency have given $4.9 million to Lockheed Martin to develop an image recognition system that will be used to scan satellite images and photographs for familiar objects. Called Object Recognition via Brain-Inspired Technology (ORBIT), the system will fuse commercial airborne EO and LIDAR sensor data into a three-dimensional, photo-realistic model of the landscape. The brains of the system, so to speak, will be Numenta's Hierarchical Temporal Memory technology, modeled on the technology growing inside human heads. The system is expected to increase image analysts' productivity by 100 times."
to start SkyNet.
How we know is more important than what we know.
...and just get someone to fly around in a jet doing this? Last time I checked the average person had a brain...why do we need to spend so much cash to make a new one!
Obligatory blog plug: http://www.caseybanner.ca/
Through all the market-speak, are they spending this money on essentially (well trained, possibly overtrained) neural network models? Or is there something more to this besides 20 year old technology packaged for the military?
it doesn't half sound like some kind of fancy porn detector...
Max.
This should come in handy when we're searching for the last glaciers in decade or two
Its not the years, its the mileage
While loaded with buzzwords, this really involves nothing that's really new. The HTM is just a rehash of Adaptive Resonance Theory .ps.gz file).
And applications like this aren't exactly new (this link downloads a
Although it is certainly a major engineering challenge to get this type of classification to work over multiple modalities of data in any coherent way, as far as I can tell this project doesn't represent any breakthrough in approach or capability.
Like any human brain, it seems mesmerized by the Space Needle and of course the British Penis building.
Am I the only one wondering about the 100X productivity boost? I don't see it. Rather than using current technology to photograph and map a certain area, say the European Alps, the workers now have to add several input streams and correlate them in 3D. For the people actually producing maps today, this sounds like it will take much longer than before to produce a final product. I don't think it takes all that long to make a 2D map from a collection of 2D images.
Up to now, the hard parts are ensuring consistancy in the initial images and blending lines between separate images. With this process it involves calculating heights of mountains, depths of river valleys, and even water and ice features as well. I doubt the computers will do this as some magic algorithm, it will be human map-makers performing the exhaustive tasks. The computers will just be there to give them more data and 'guesstimate' what it will look like.
If you think I voted for Trump because of this post, you're wrong. I voted for Dr. Jill Stein of the Green Party. Again.
Yep, I knew it. These guys want to know everything, I can just imagine what kind of black-deals they can cut with major corporations for competitive advantage, using the data being fed by these satellites to determine patterns in human behaviour and using that for strategic investment. And that is only the tip of the iceberg...
Tinfoil hat you say? One only has to look at history, Alexander thought himself a god (or wanted to be one) and man is obsessed with improving his power to dominate and control both peaceful and hostile populaces, the truth of the matter is, why let the future happen to you when you can start to predict it, and thereby shape it?
That is what the power mongers of this world want, is some modicum of ability to guide and shape history in their favor. And if you were at the top, among competitors that may beat you to it... you'd want it too.
Urban Reasoning and Geospatial Exploitation Technology (URGENT)
The Urban Reasoning and Geospatial Exploitation Technology (URGENT) program is will develop a 3D urban object recognition and exploitation system that enables advanced mission planning and situation analysis capabilities for the warfighter operating in urban environments.
The recognition of targets in urban environments poses unique operational challenges for the warfighter. Historically, target recognition has focused on conventional military objects, with particular emphasis on military vehicles such as tanks and armored personnel carriers. In many cases, these threats exhibit unique signatures and are relatively geographically isolated from densely populated areas. The same cannot be said of today's asymmetric threats, which are embedded in urban areas, thereby forcing U.S. Forces to engage enemy combatants in cities with large civilian populations. Under these conditions, even the most common urban objects can have tactical significance: trash cans can contain improvised explosive devices, doors can conceal snipers, jersey barriers can block troop ingress, roof tops can become landing zones, and so on. Today's urban missions involve analyzing a multitude of urban objects in the area of regard. As military operations in urban regions have grown, the need to identify urban objects has become an important requirement for the military. Understanding the locations, shapes, and classifications of objects is needed for a broad range of pressing urban mission planning analytical queries (e.g., finding all roof top landing zones on three story buildings clear of vertical obstructions and verifying ingress routes with maximum cover for ground troops). In addition, it will enable automated time-sensitive situation analysis (e.g., alerting for vehicles found on a road shoulder after dark and estimating damage to a building exterior after an explosion) that will make a significant positive impact on urban operations.
Phase 1 of the URGENT program is developing techniques for the rapid exploitation of EO and LIDAR sensor data at the city scale to recognize urban objects down to the soldier scale. URGENT is applying image processing technology to geospatially registered 2D/3D data collected from airborne and terrestrial sources, yielding precise annotations for the objects in an urban area.
Phase 2 of the URGENT program will develop a 3D reasoning engine to query over object shapes, locations, and classifications for rapid urban mission planning, mission rehearsal, and situation analysis. Phase 3 will focus on the integration and transition of the URGENT system to the National Geospatial-Intelligence Agency (NGA).
How we know is more important than what we know.
I don't want SkyNet. I just want a friend to talk to...
The good news is that this is all math! There's no need to believe anything one way or another! Sorta exciting huh? You can go and examine all the ART algorithms (I linked wikipedia because it has the PDFs linked.. did you notice? But here's Grossberg's homepage, just in case), and you can go read about HTM. According to Hawkins, HTM has some magical, er I mean, proprietary, component that separates it from ART. I've seen Hawkins speak... in fact, I saw him speak at BU with Steve Grossberg in the audience. He amused the audience by showing a demo that was completely indistinguishable from an ART1 implementation that takes about half an hour to program, and most of the people present had done themselves.
He then failed to answer any substantive questions (including Steve asking him how his model differed from ART), referring us all to online videos of his lectures. I personally asked about how he could reconcile this article with his predictions.. which assume a cortical hierarchy based on 'distance' (in synapses) from primary sensory cortices, rather than examining the relative lamination of various cortices. I notice since then the wikipedia article "On Intelligence" has had its 'experimental prediction' claims toned down quite a bit.
As it happens in terms of books though, Grossberg has written several and has a ton of peer reviewed articles on this very subject. Hawkins to my knowledge doesn't have a single peer-reviewed article on HTM or anything related.
"He who appeals to authority when there is a difference of opinion works with his memory rather than his reason."
Please consider this post about Jeff Hawkins' history of navel-gazing idiocy in the field of neural networks.
You worship engineers? Why this one in any case? Even if he is a good engineer, that doesn't make him a good scientist (incidentally, he's not). Maybe you're not an expert in neural networks and are deferring to someone who is, at least plausibly, an expert. But you just illustrated that you're no better a thinker than the people you probably decry day-in, day-out; you're just one of their number who reads slashdot and gets an undeserved superiority complex. Did you even consider that "some guy on slashdot" provided a link to an established area of theory that is (if you investigated even a bit) AT LEAST similar to Hawkins' stuff?
Jeff Hawkins is not even noteworthy in the fields of neural networks or AI in general. He's got a lot of bluster and a lot of money though, and he's using that money to get the popular (in geek circles) press gain the "prestige" that can only be bought through useful, novel research in academia. He's not an expert, and he's probably a charlatan for claiming to be a cutting-edge researcher merely commercializing "his" (ahem) "new" (ahem) breakthroughs.
His disingenuous pretense is what gets me about him. Your unthinking deference to "what you read in some book" is what gets me. Please, if you're interested in neural nets (you seem to be), read some of the cornerstones in the field, read the academic literature, and stay away from Numenta and Hawkins.
Feature extraction, such as buildings and trees is already being done. In fact, the dot clouds produced by LIDAR provide enough information about trees that some research is focusing on ways to automate the identification of species of individual trees, and replicating that across an entire forest. But I digress... I, for one, welcome our Numenta powered, LIDAR scanning, ORBIT-Lords.
Authority questions you. Return the favor.
I was wondering what the helmet icon was about, but then I realised this is DAAAAARPA.
Jeff has proposed a theory of how the cerebral cortex works, which is not in itself unusual. There's lots of people who have proposed outlandish solutions to the various problems posed by AI. They're usually labelled crackpots.
Jeff's proposal is different in that he has actual working code based on his theories which do a pretty good job of recognizing things. And he bases his theories on his interpretation of how the cortex actually works - from neuropsychology studies. Not from suppositions founded on math and creative thinking.
In my opinion, Jeff's stuff is correct but incomplete. His software makes a great recognizer, but the human brain contains other parts that his software doesn't address. His software has no provisions for goal seeking, for instance (the reticular activation system), or attaching an "emotion" to a memory as a way of indicating the advisability of repeating it (hippocampus, perhaps).
His system has the capability to predict the future outcome of current actions, but no method of rooting around in the various possible outcomes in order to choose a course of action which is beneficial to the organism's goals. It can't think ahead, or even identify the need to.
His work is also unclear about how the system generates outputs. Having chosen a course of action, speaking a thought perhaps, there is no clear description of how the system "unrecognizes" the goal into its base output components - the speaking or writing motions which would cause the output to be rendered.
His work is alluring in that it seems to reflect current models of brain physiology and function, and even anecdotally with my own inner workings and those of other people.
You shouldn't count him out with such a small wave of the hand. Certainly not without more direct reasoning, comparison, or critique.
.... these being the brains that get abducted by aliens, and see images of the virgin mary in slices of toast?
Good luck guys.
Old COBOL programmers never die. They just code in C.
You are totally correct on the literature.. I was posting quickly ;) Rao & Ballard in particular I think influence Hawkins works (not to mention that Rao was one of the people who convinced me to go into neuroscience.. but I digress). The problem with Lee & Mumford is that it's now been known for quite a while that V1 receptive fields are not static, but dynamic in really cool ways.. Check out Ohzawa's videos for example.
welcome our new ORBIT see-it-all overlords.
just a few hours after I wiped out Numenta's software from my HDD
Anyone else think for a moment that Shrub got his own Slashdot topic or just us Doonesbury fans?
Kwisatz Haderach
Sell the spice to CHOAM
This Mahdi took Shaddam's Throne
According to TFS:
The brains of the system [...] will be [...] modeled on the technology growing inside human heads.So, dude... the technology is growing inside your head now.
Been nice knowing you.
Last I checked, there wasn't any technology growing inside my head. Am I living in the wrong 2007 or something?
Doesn't it make you feel good to know that our freedoms are protected by politicans, lawyers and journalists.
What exactly is the point of all this?
Although the concept of using LIDAR to create an extremely detailed topographical map is certainly a neat (and useful) thing to do -- military and non-military applications alike, I question exactly how the AI engine is going to come into play.
This sort of system would make sense if you were scanning for subs and stealth aircraft from space -- the sort of thing that has a regular shape. But as to our current military situation, how the heck are you going to correctly deduce the presence if an IED? An IED is exactly that -- improvised. There's no standard model that can easily be spotted in a photograph. It doesn't even need to be particularly large or conspicuous to cause some serious harm.
Couldn't you also fool the system by simply throwing a tarp over whatever it is you're trying to hide? If you wanted to hide some sort of armored vehicle, couldn't the system be easily defeated by covering it in camo-netting? It'd look like a shrubbery when viewed vertically.
The way we fight wars is completely different than 60 years ago. Any power with a large, modern, and organized military able to fight a "conventional" war already has nukes. Likewise, as our progress in Iraq is currently demonstrating, a conventional military performs miserably against guerilla tactics (throwback to the American Revolution?). Mind you, there were all sorts of other mistakes made with Iraq that could have potentially prevented this sort of warfare from erupting (the Powell Doctrine comes to mind), but the fact is that we're now forced to deal with it, and when the people you're trying to protect are also the ones trying to kill you, your entire military strategy is utterly useless.
All in all, this sounds like a very expensive solution looking for a problem. I appreciate the research done by DARPA, but this is a waste of their time.
-- If you try to fail and succeed, which have you done? - Uli's moose
way to go darpa!
www.emailshadow.com