Slashdot Mirror
DARPA's Cortically-Coupled Computer Vision System
BluePariah writes "Wired News has an article on a 'cortically coupled computer vision' system being developed at Columbia University and funded by the ever-curious folks at DARPA. Essentially, it uses the extremely powerful visual recognition ability of the human brain and couples it with a computer's raw processing power to allow a user wearing an EEG cap to filter through scores of digital images at high-speed and pick out something of interest. This has applications in military intelligence, face-recognition, anti-terrorism, and hunting down replicants."
Blip-Verts!
The TV networks will love this!
Why, oh why, didn't I take the Blue Pill?
[...] allow a user wearing an EEG cap to filter through scores of digital images at high-speed and pick out something of interest.
hi-speed pr0n!
Stop Computers/Cars Analogies on S
So DARPA's invented something else now. How long before Al Gore goes on CNN to claim he invented this all by himself as well?
Say it with me now... Porn!
GetOuttaMySpace - The Anti-Social Network
*Banana clip for your face sold separately
Where does the school board find them and why do they keep sending them to ME?
Suprise: I thought that the human visual systems is way superior to the existing computational image-processing systems. But I guess this technology switch directions as well, switching the roles and using human brain as co-processors in surveillance and security applications. Any volunteers for this?
seems like a first case of positive usage of subliminal messages. I wonder though, who would accept his brain to be fried in such a way.
cut this signatures madness. stop reading them now!
"and hunting down replicants."
Hey, they are only guilty of DNA copyright infringement! It's not like it's an actual crime, bud!
Where were you when the voynix came?
Not only do you not have to press the button but you can look at the data ten-times as fast. Imagine this scenario: You're a Homeland Security Goon at the airport and the boys in the NSA have provided you with the face of a terrorist that may be walking around the airport. You memorize the picture of the guy, put on your nifty EEG space helmet, and tap into the face recognition system camera database at the airport. You then sift through thousands of photos in mere minutes. Human recognition works FASTER than human consciousness and therefore can identify the images before you even 'know' it. The EEG can detect the signals of your brain recognizing images and when it gets a 'hit', it dumps that image them into a cache for closer review at a later time. Think about it... banks of people in a windowless office with EEG helmets on pouring over pictures from every corner of the globe looking for whomever. Fascinating and scary at the same time...
Ooo man the floppy drive is broken. No wait. The computer is just upside down.
Replicants is a reference to Blade Runner. A move by Ridley Scott.
The IMDB link: http://www.imdb.com/title/tt0083658/
The move is based on the work of Phillip K. Dick. It also stars Harrison Ford in his least favorite role.
So, basically completely the opposite to the /. description, to whit:
.I picked that up within 5 seconds of clicking the link. Sort it out, editors.
Meta will eat itself
How soon until we get distributed-image-glancing teams together, racking up spare brain cycles for high scores and bragging rights?
Slashdot Burying Stories About Slashdot Media Owned
IXO (DARPA'S Information eXploitation Office) just made awards for their VACE (Video Analysis and Content Extraction) BAA and this sounds a lot like some of the technologies they were trying to develop through that program. I'll have to do more digging, the article itself is somewhat suspect (some jackass with a Ph.D. in *transport systems* flaunting his ignorance of computer vision isn't exactly a good source to quote). I particularly like the bit about "They are limited in their ability to recognize suspicious activities or events." Turns out that he hasn't read Grimson and Stauffer's (fellow MIT alums) papers. Or, you know, about 20-30% of the computer vision literature.
Shees, next it's going to be 2001, The Time Machine, and Ice Pirates
Wait, wrong argument
- where will it end - check
- think of the children - skip
- work of terrorists - skip
- violation of civil rights - check ^H^H - skip
- end of civilization - change civ to cult - check
OK, back....to allow this affrontery to continue will undoubtably lead to the end of Western Culture as we know it, for without due veneration of our classical arts, we shall indeed be doomed to an eternity of Jerry Springer and Teletubbies. Oh the humanity of it all.
Even better, we could breed people specifically to serve this purpose. We could have a whole class of people created in test tubes, deprived of meaningful human contact and trained just to look at thousands of images per minute, all day every day. With banks and banks of these people, we could create a human powered "supercomputer" capable of identifying any person on earth in mere seconds!
Note: This is currently a fictional scenario, but in one hundred years when this is actually going on, someone will stumble upon this post and realize how very forward-thinking I was...
-Arthur
Cave ne ante ullas catapultas ambules
Don't tell the MPAA! By feeding digital images directly into the brain of the viewer, they've finally managed to get rid of that nasty analog hole that pirates are always exploiting.
Support Right To Repair Legislation.
One of the basic tasks our visual system is much, much better at executing than computers is visual search. The basic 'experiment' is that you are asked either a question like "Is there a red car in this picture?" (natural images) or "Are all the lines the same orientation?" (more traditional psychophysics). Then images are displayed, and our response time is recorded. Early experiments in the visual search paradigm appeared to show that there was two classes of search stimuli: those that 'pop-out' and those that require incremental search. The difference is that in pop-out conditions, increasing the number of elements in the image does not increase search time, while in incremental it does at XXms/element... and generally it takes about twice as long for us to respond if there is no positive element.
One main theory on how our brain does this, Feature Integration Theory by Anne Treisman (or similar but more recent, Guided Search by Jeremy Wolfe), which many computer vision algorithms try to copy, asserts that there are various feature maps for certain quantities like color, orientation, depth, spatial scale, etc. These are combined into a saliency map which is a weighted average of the feature maps. Things pop-out when the target has high salience compare to the background, for example it's easy to find the red T in a background of blue T's, but not so easy to find the red L in a background of red T's and blue L's.
Now, it appears from the article, and what little they say on the Lab webpage, that they are trying to measure EEG responses (which are quite crude) during rapid serial search tasks, in order to prime a computer vision object recognition system, which is then only run on those images human's appear to find sufficintly salient when they see them. This saves the time of a person actually having to search and make a decision about an image, while utilizing the visual systems incredibly powerful early 'pre-attentive' form & object binding resources.
If there is a sufficiently high signal from the EEG to do that after say, 100ms display times, then I think this could be useful for certain types of search task. However, due to the time courses present in most visual search experiments, the fact that it's not totally apparent how efficient certain parts of our saliency system actually are (check our Jeremy Wolfe's reviews for more data), I'm totally unconvinced that this type of system will give you a sufficent signal to noise ratio to be worth using for anything. This is especially true because of another perceptual phenomenon in search, which is that your error rate basically shoots up exponentially as the probability of a positive goes down. This is to say, in an experiment where a normal observer would have a 99% accuracy rate with 50% of the images containing the target, this drops to 60% accuracy for 10% target positive, and only 30% accuracy at 1% target positive (numbers fudged, but ballpark, since I'm too lazy to look them up). If this has its roots in insufficient priming in early vision, for example, then this entire scheme flops just as badly as using a human for tasks like finding the bomb in the x-ray image of the suitcase... and we haven't even started to get into issues of the person not actually looking at the image because they're bored, etc.
As it is, DARPA is spending a mere 758k, which is chump change for them, and there's a decent chance that it'll work in certain specific but useful circumstances which may warrant the research.