Posted by ljmacphee on April 28, 2008 under artificial intelligence in the news, computer vision | 
New insight into how brains process visual information is a double edged sword. It will make for much better vision engines but with that will come the failure of our most popular human test at the moment — captcha.
Using a fly, whose brain is heavily coded for visual information, Nemenman and his colleagues were able to show information is passed along the spikes in the fly’s brain neurons.
. . .
Nemenman and his colleagues’ research is significant because it re-examines fundamental assumptions that became the basis of neuromimetic approaches to artificial intelligence, such as artificial neural networks. These assumptions have developed networks based on reacting to a number of impulses within a given time period rather than the precise timing of those impulses.
“This may be one of the main reasons why artificial neural networks do not perform anywhere comparable to a mammalian visual brain,” said Nemenman, who is a member of Los Alamos’ Computer, Computational and Statistical Sciences Division. “In fact, the National Science Foundation has recognized the importance of this distinction and has recently funded a project, led by Garrett Kenyon of the Laboratory’s Physics Division, to enable creation of large, next-generation neural networks.”
New understanding of neural function in the design of computers could assist in analyses of satellite images and facial-pattern recognition in high-security environments, and could help solve other national and global security problems. [ read more Language of a fly proves surprising ]
Papers:
PLoS: Neural Coding of Natural Stimuli: Information at Sub-Millisecond Resolution
More information:
Is Captcha’s moment passing?
Posted by ljmacphee on April 24, 2008 under artificial intelligence in the news |
Totally cool and totally scary. This algorithm finds hidden social networks no matter how small. This may turn out to be an excellent resource against terrorist networks. Currently the algorithm has and is being used to detect genetic networks. The algorithm was inspired by stegography but can be applied to any network of information.
Human diseases and social networks would seem to have little in common. However, at the crux of these two lies a network, communities within the network, and farther even, substructures of the communities. In a recent paper in Physical Review E 77:016104 (2008), Weixiong Zhang, Ph.D., Washington University associate professor of computer science and engineering and of genetics, and his Ph.D. student, Jianhua Ruan, published an algorithm, a recipe of computer instructions, to automatically discover communities and their subtle structures in various networks.
Many complex systems can be represented as networks, Zhang said, including the genetic networks he studies, social networks and the Internet. The community structure of networks features a natural division of the network where the vertices in each subnetwork are highly involved with each other, though connected less strongly with the rest of the network. Communities are relatively independent of one another structurally, but it is thought that each community may correspond to a fundamental functional unit. A community in a genetic network usually contains genes with similar functions, just as a community on the World Wide Web often corresponds to web pages on similar topics.
All Zhang and Ruan need are data. Their algorithm is more scalable than existing algorithms and can detect communities at a finer scale and with a higher accuracy than similar algorithms. The impact of having such a computational biology tool is in genomics, where researchers may be better able to identify and understand communities of genes and their networks as well as how they cooperate in causing diseases, such as sepsis, virus infections, cancer and Alzheimer’s disease. [ read more Algorithm finds the network -- for genes or the internet ]
More information:
Jianhua Ruan’s Homepage
Weixiong Zhang’s Homepage ( links to several of his programs on homepage )
Nature paper describes technique for extracting hierarchical structure of networks
Papers:
In Search of the Biological Significance of Modular Structures in Protein Networks
Posted by ljmacphee on April 21, 2008 under artificial intelligence in the news |
Some UA Huntsville researchers who specialize in statistics are finding patterns in asymmetric threats to the US and US troops. While these attacks seem random some patterns are emerging.
While these patterns do not give specific information as to what will be attacked, when and how, it does give probabilities of likely targets, types of attacks and time frames. Greater resources can then be applied to those areas over that time frame.
But the question is; is there really a pattern? We humans can find patterns in almost any sufficiently large database of information. It is also well known any well published attack will attract copy cats.
If it turns out that there are patterns to these attacks what’s to stop the terrorists from using a random number generator to plan future attacks?
Since the Sept. 11 attacks we’ve had several lists of cities likely to be hit by terrorists. Not of which has been attacked. And we’ve had several predictions of types of terrorist attacks to come. So far all they have predicted is what else will need to be dismantled before you can board a plane.
If we hope to anticipate terrorism perhaps we need to look at something besides statistics and predictive markets?
More information:
Computer Models to Provide Better Intelligence for Army
RAND Tries to Model Risks of Terrorist Attacks
The statistics of fear
Eigenbehaviors
Prediction Markets are hot but here’s why they can be so wrong
Posted by ljmacphee on April 17, 2008 under artificial intelligence in the news, computer vision |
Ah, but do we really want our computers to understand us? Anybody remember ‘Clippy’?
Computer: “You seem depressed today, should I Google Dr Kevorkian for you?”
Or will the clerks at the local retail store start wearing cameras with emotion recognizing software? A bit of customer understanding by the help would go a long way in many a business.
Researchers at the Department of Artificial Intelligence (DIA) of the Universidad Politécnica de Madrid’s School of Computing (FIUPM) have, in conjunction with Madrid’s Universidad Rey Juan Carlos, developed an algorithm that is capable of processing 30 images per second to recognize a person’s facial expressions in real time and categorize them as one of six prototype expressions: anger, disgust, fear, happiness, sadness and surprise.
Applying the facial expression recognition algorithm, the developed prototype is capable of processing a sequence of frontal images of moving faces and recognizing the person’s facial expression. The software can be applied to video sequences in realistic situations and can identify the facial expression of a person seated in front of a computer screen. Although still only a prototype, the software is capable of working on a desktop computer or even on a laptop. [ read more Facial Expression Recognition Software ]
How ever emotion recognition software gets used in the future this software is bound to be fun.
More information:
Video of software in action
Papers:
Jose Miguel papers ( $ )
See also:
Software recognizes anxiety in people
Posted by ljmacphee on April 14, 2008 under artificial intelligence in the news, robotics |
I don’t know how I missed this story last year. Flying robots that repair satellites are cool. Maybe in time we’ll have robot mechanics patrolling the highways and fixing disabled vehicles?
Boeing Orbital Express system is a DARPA project hoping to demonstrate fully autonomous on orbit satellite servicing capabilities. Which in plain English means robot satellites who service and repair other satellites. These satellite mechanics will fly as high as 7 km.
Last year Boeing successfully demonstrated refueling and battery replacement using autonomous robots. Using robot satellites to refuel and do maintenance on other satellites will greatly drop the cost of using satellites.
Two Defense Advanced Research Projects Agency (DARPA) Orbital Express
demonstrator spacecraft were launched yesterday at 10:10 p.m. EST, from Space Launch
Complex 41, Cape Canaveral Air Force Station, Fla., aboard an Atlas V 401 booster. The launch
was sponsored by the Department of Defense Space Test Program (STP), as part of the STP-1
mission, which includes, in addition to Orbital Express, the STPSat-1 spacecraft and three other
microsatellites.
Orbital Express consists of a next generation serviceable “client” satellite (NextSat) and a
prototype servicing spacecraft (Autonomous Space Transport Robotic Orbiter or ASTRO). They
were deployed together into a circular, 492-kilometer, low earth orbit with an inclination of 46
degrees. They will spend the next three months in orbit, demonstrating for the first time fully
autonomous rendezvous and capture of client spacecraft, satellite-to-satellite refueling, and
replacement of battery and flight-computer orbital replacement units.
More information:
Boeing Orbital Express
The REal R2D2: ASTRO Takes on Space Rogues
Darpa: Orbital Express ( has videos and photos )