Emerging Technology Trends

Computer scientists in California have built a mathematical theory of surprise, working from first principles of probability theory applied to a digital environment, according to this news release from the University of Southern California (USC). And the results of experiments recording eye movements of volunteers watching video seem to confirm it. Beyond vision applications, this new Bayesian theory of surprise could lead to new developments in data mining, as it can in principle be applied to any type of data, including visual, auditory or text."

This new mathematical theory of surprise has been developed by Laurent Itti, of the USC's Viterbi School of Engineering, his colleagues at his lab, and by Pierre Baldi, of the University of California Irvine's Institute for Genomics and Bioinformatics.

Before looking at their theory, here are some key definitions given by the computer engineers.

[By analyzing streams of electronic data making up a video image,] researchers can isolate stimuli with visual attributes that are unique in the mix by breaking down the signal into "feature channels," each describing a particular attribute (i.e,, color) in the mix. Such features are called "salient."

A parallel analysis performs similar operations, but does so over time, not space, looking for new elements suddenly appearing. This approach is said to model "novelty."

Finally, an analysis can be done purely in terms of Shannon's original equations, which can measure the level of organization or detail found in the data flow, its entropy.

And now, let's look at their theory.

Their theory boldly proposes to make just such predictions, working from probability theory as well as digital principles. The probability theory involved is that known as "Bayesian," which amounts to a way of structuring events observed over time in the past into predictions about the future.

And they put their theory at work.

The next step is to use this theory to analyze a video stream to describe what are the streams most "surprising," features. Finally, having performed this analysis, they checked it by watching the eye movements observers watching the images, to see if the eyes followed the measure of surprise.

On the figure below, you can see: "(a) Sample eye movement traces from four observers (squares denote saccade endpoints): (b) Our data exhibits high inter-individual overlap, shown here with the locations where one human saccade endpoint was nearby one (white squares), two (cyan squares), or all three (black squares) other humans; (c) A metric where the master map was created from the three eye movement traces other than that being tested yields an upper-bound KL score, computed by comparing the histograms of metric values at human (narrow blue bars) and random (wider green bars) saccade targets. (Credit for image and caption: USC). (more…)

Back in August 2003, I told you that spray-on nanocomputers were coming. Last week, the Discovery Channel gave us an update on this project of the Speckled Computing consortium, a collaboration between researchers from several universities in the U.K. They are building the Specknet, a network of small computers (5 x 5 x 5 mm), which also contain sensors and can communicate wirelessly with the other specks. But it will probably take ten years before such dynamic sensors can be used in real applications, like detecting structural failures in aircraft wings or helping victims from strokes.

Here are the opening paragraphs from the Discovery News article.

Grain-sized semiconductors could one day be sprayed onto surfaces like paint onto walls to give computers access to places previously out of reach.

The so-called Specknet combines sensing, computer processing and wireless communication to link the physical and digital world in a kind of computational aura.

The Specknet is currently under simulation at the Speckled Computing consortium. Its director, D.K. Arvind, who also works at the Institute for Computing Systems Architecture at the University of Edinburgh, tells us more about the tiny computers which compose this network.

"Because they are so small, you can extend computing and sensing to areas that couldn't be reached before."

And he also gives us the reason why these sensors have some processing capabilities.

"The rule of thumb is that it is ten times more expensive to process a byte of data remotely than process it locally."

[So] in a Specknet, each sensor has its own processor, about two kilobytes of memory, and a program that gives it the ability to extract information from the environment, collaborate with other local specks, and act on the data gathered.

Below is a picture of a mockup of one of these Specknet sensors, called the 5CubeOTS, a 5 x 5 x 5 mm sensor node which will be built from off-the-shelf components (Credit: Speckled Computing consortium).

And below is another Specknet element, dubbed ProSpeckz for "Programmable Specks over Zigbee radio". (more…)

You've probably read that an hydrogen-based economy will emerge soon — or in twenty years. But first, we need to find efficient ways to store hydrogen fuel. Now, chemists from the University of Michigan (U-M) have developed new materials made of rigid plastics that are at the same time lightweight, robust and economical. These new materials, called covalent organic frameworks (COFs), are made of light elements, such as hydrogen, boron, carbon, nitrogen and oxygen, and can be assembled to form predictable crystal structures — something that never had been done with rigid plastics. So far, these materials are only available in labs.

Before entering technical details, it's interesting to discover why a young scientist wants to enter a specific research field. Omar Yaghi is a professor of chemistry at U-M involved in this COFs work. Here is a link to a recent interview with him where he tells us about his motivations.

It was the aesthetic aspects of chemistry, rather than the utilitarian, that first lured him into the field, the 40-year-old professor recalls. But his urge to understand the molecules he found so attractive eventually led him to develop materials with very practical applications: storage of hydrogen, methane and other economically and environmentally important gases.

"I didn't set out to solve any big problems," Yaghi admits. "I was drawn to chemistry mostly because of the beauty of the molecules." Browsing through a college catalog one day, he came across pictures of cage-like chemical structures similar to buckyballs. Yaghi "fell in love" with the structures and decided on the spot that he wanted to learn how to design and make chemical structures of various shapes.

Now, it's time to listen to Adrien Côté, who is a member of Yaghi's laboratory, to discover what these researchers have created.

"Normally, rigid plastics are synthesized by rapid reactions that randomly cross-link polymers," said postdoctoral fellow Adrien Côté. "Just as in anything you might do, if you do it really fast, it can get disorganized." For that reason, the exact internal structures of such materials are poorly understood, making it difficult to predict their properties. But Côté and colleagues tweaked reaction conditions to slow down the process, allowing the materials to crystallize in an organized fashion instead of assembling helter skelter.

Below is a picture showing the crystalline sheets produced in covalent organic frameworks (COFs) (Credit: Adrien Côté, U-M). And here is a link to a larger version ( 3,400 x 4,400 pixels, 1.45 MB).

And below is a scanning electron microscopy (SEM) image of "a COF-5 revealing the piles of deformed hexagonal plates" (Credit: U-M). (more…)

For almost twenty years, the iris recognition research field has been hampered because of a broad patent covering it. As this patent recently expired, many teams around the world are again working on new technologies in this field. Iris recognition is in fact seen as the most accurate biometric recognition technology because no two irises are identical. And researchers at the University of Bath in England have developed new computer algorithms which are 100 per cent accurate in initial trials. Now the researchers are putting online a database of 16,000 iris images collected mainly from students. The source code is also available if you want to further improve the algorithms.

Before going further, let's go back in time to understand why this research field was almost inactive for twenty years. Life Style Extra tells us the story.

Looking into a camera to confirm your identity would now be routine and - were it not for the US firm's virtual monopoly of the technology - it would already be in use at cashpoints and passport control. Its backers say it could reduce fraud and illegal immigration.

Iridian Technologies, based in New Jersey, patented the system of identifying people using the coloured part of the human eye in the mid 80s and other scientists have had to pay tens of thousands of American dollars to do any research in the field, thus hampering competition.

But the patent expired in the US earlier this year and expires in the rest of the world in February 2006.

Now, it's time to return to 2005 at the University of Bath.

Engineers are currently road-testing their technology using a specially-constructed database containing thousands of iris images collected from students and colleagues at the University.

By making this database available to other research groups, the researchers hope to encourage more advances in iris recognition and overcome some of the restrictions caused by a generic patent (recently expired) which has limited innovation for the last two decades.

"Our new algorithm does the same job as the one used by almost all of the commercially available iris recognition systems, it just does it better," said Professor Don Monro from the University’s Department of Electronic and Electrical Engineering.

Below is a picture showing how an iris picture is shot and rendered on a computer screen before being analyzed (Credit: Smart Sensors Ltd.).

And below is an illustration of the iris image normalization process (Credit: Smart Sensors Ltd.). (more…)

According to InfoWorld TechWatch, new scanners using a technology developed at the U.S. DOE's Pacific Northwest National Lab are about to appear at airports. The scanning process lasts only 1.5 second and produces on a computer screen a holographic image of the individual’s body and any objects he or she is hiding. But don't be afraid: there is a privacy mode, which can blur some parts of your body, such as your chest or rear. These scanners will be deployed not only in airports, but also in prisons, military caps and at vorder crossings.

Here are the opening paragraphs from the InfoWorld article called "Total Recall body scanner becomes reality."

Replete with a privacy algorithm to hide "the important [body] parts," no not your brain, the vice president of operations of SafeView, Karen Meyer, gave me the rundown on the Scout 100 and Scout 360 people scanners.

This baby is much closer to the scanner in the Arnold Schwarzenegger movie Total Recall than the current generation of two-dimensional x-ray scanners.

Below is a picture of the Safescout 360 model (Credit: SafeView).

Here is link to an article from Congressional Quarterly Homeland Security, "Fabulous: When Fashion Sense Meets Security Needs," which gives more details about these scanners.

The SafeScout portal looks like a small revolving door, minus the actual doors. Inside there are speakers and video cameras, allowing security personnel to communicate with the individual from a remote location.

The airline passenger, border crosser or prison entrant steps inside and, in the case of one model, the entrance and exit doors close automatically. A pair of vertical bars positioned between two layers of glass swing 360 degrees and then stop.

That was the 1.5-second scan. A second later, a holographic image of the individual’s body and any objects he or she is hiding appears on a computer screen.

InfoWorld adds that these scanners allow to scan 400 people per hour.

For more information about these scanners, here are two links to the datasheets of the SafeScout 100™ and the SafeScout 360™.

And if you want more details about the active millimeter wave scanning technology developed by the U.S. DOE's Pacific Northwest National Lab, here is a short news release from the lab, Screening for security.

You also should read a technical paper called "Concealed explosive detection on personnel using a wideband holographic millimeter-wave imaging system." Here is a link to this document (PDF format, 11 pages, 736 KB).

Sources: Ephraim Schwartz, InfoWorld TechWatch, November 22, 2005; Zack Phillips, Congressional Quarterly Homeland Security, June 17, 2005; and various web sites

You'll find related stories by following the links below.

• Military Applications
• Police
• Security
• Technology
• Travel
• Vision and Visualization

Yesterday, I told you about researchers who visualized the effects of stress on the human brain. Let's focus today on the activities of our brain during and after hypnosis, with the help of the New York Times (Free registration, but permanent link). For example, you'll discover that "new experiments, which used brain imaging, found that people who were hypnotized 'saw' colors where there were none. This particular research has also been covered a few months ago by Nature and by Scientific American.

Let's start this overview by an introduction from Scientific American.

The word "hypnosis" tends to conjure up images of subjects partaking in silly activities they might not otherwise agree to. But over the past few decades, scientific study of hypnosis has begun to identify how the approach can work to alter processes such as memory and pain perception. According to a new report, hypnotic suggestions regulate activity in certain regions of the brain and can help it manage cognitive conflicts.

This new report was co-authored by Amir Raz, an assistant professor of clinical neuroscience at Columbia University, who works at the Brain Imaging Lab. Raz wanted to know how hypnosis could be used to reduce these cognitive conflicts and based the study on the Stroop test. The New York Times explains what is this test.

The probe, called the Stroop test, presents words in block letters in the colors red, blue, green and yellow. The subject has to press a button identifying the color of the letters. The difficulty is that sometimes the word RED is colored green. Or the word YELLOW is colored blue.

For people who are literate, reading is so deeply ingrained that it invariably takes them a little bit longer to override the automatic reading of a word like RED and press a button that says green. This is called the Stroop effect.

You can experiment this effect yourself. Below is an example taken from a series of optical illusions on display on the At-Bristol website. And here is a link to the page about the Stroop effect.

So how did Raz use this test? (more…)

According to this news release from the University of Pennsylvania School of Medicine, researchers have for the first time "visualized the effects of everyday psychological stress in a healthy human brain." They used the fMRI (functional magnetic resonance imaging) technique and found that there is a strong link between psychological stress and negative emotions. This research effort may help physicians to better diagnose and treat the effects of chronic stress.

Here is how the researchers did their study.

In the Penn study, researchers induced stress on healthy subjects by asking them to quickly tackle challenging mental exercises while being monitored for performance. During the fMRI scans, the researchers also recorded subjects' emotional responses — such as stress, anxiety, and frustration — and measured the corresponding changes in stress hormone and heart rate. Many subjects described themselves as being "flustered, distracted, rushed and upset" by the stress task.

The background of the image below "shows the mean cerebral blood flow of all the subjects undergoing the stress tasks (here, mental arithmetic) acquired using the continuous arterial spin labeling (CASL) technique. The foreground shows the detected activation in the right prefrontal cortex — an area long associated with anxiety and depression — when the subjects are under stress." (Credit: University of Pennsylvania School of Medicine).

And here is a link to a larger version of this image.

And here are two miniature images reflecting the brain's activity when the subject tells a lie (left) or the truth (right) (Credit: University of Pennsylvania School of Medicine). Here is a link to a poster from which this image was extracted.

Before going further, you also should watch one of the videos available from the Virtual fMRI page. The subjects need to spend about one hour in the MRI coil, without moving, and subjected to various stimuli. This should be largely enough to increase my level of stress…

But let's come back to the results of this study, which were driven by John A. Detre [,Associate Professor of Neurology and Radiology] and by Jiongjiong Wang [, Assistant Professor of Radiology.]

The results showed increased cerebral blood-flow during the "stress test" in the right anterior portion of the brain (prefrontal cortex) — an area long associated with anxiety and depression. More interestingly, the increased cerebral blood-flow persisted even when the testing was complete.

These results suggest a strong link between psychological stress and negative emotions. On the other hand, the prefrontal cortex is also associated with the ability to perform executive functions — such as working memory and goal-oriented behavior — that permit humans to adapt to environmental challenges and threats.

Will this study be useful for us who are all submitted to stress? Wang thinks so.

"The message from this study is that while stress may be useful in increasing focus, chronic stress could also be detrimental to mental health."

For more information, including online publications not available today, you should visit the site of the Center for Functional Neuroimaging at the University of Pennsylvania.

Sources: University of Pennsylvania School of Medicine news release, November 21, 2005; and various web sites

You'll find related stories by following the links below.

• Medicine
• Psychology
• Vision and Visualization

Current solid-state laser technologies have become vital components in telecommunications and networking equipment, but are reaching limits in speed and efficiency. But now, researchers at Stanford University have developed a new laser technology which wastes less energy than today's lasers and is faster, able to operate at rates exceeding 100 billion bps. And because up to 400,000 lasers will be able to fit within a square centimeter chip, this also means that this new technology will be cheaper to manufacture. So far, this technology is only working in labs, so it might take some time before the industry adopts it.

This new laser technology has been developed by electrical engineering Assistant Professor Jelena Vuckovic and doctoral student Hatice Altug, who is part of Vuckovic's group at the Nanoscale and Quantum Photonics Lab.

Below is a picture of Jelena Vuckovic in her lab (Credit: Stanford University). Here is a link to a larger version of this photo (3,008 x 2,000 pixels, 1.95 MB).

How does this new laser technology work?

The new laser is based on a photonic crystal, a square layer cake of indium phosphide-based material that is 300 billionths-of-a-meter (nanometers) thick and that is etched to create an array of regularly spaced, 400-nanometer-wide holes through the cake. At regular intervals among the holes are areas where no hole has been etched, called microcavities, which trap light. The filling of the layer cake is four layers of indium gallium arsenide phosphide. Each layer is called a "quantum well."

How does the laser work? When Vuckovic shines pulses of light onto the crystal, the energy that the light "pumps" into the quantum wells excites them to emit light of a desired wavelength. That light then bounces around in the microcavities and back into the wells again, setting off a chain reaction of light emission from the crystal that produces a laser beam.

And here are some details on how this new laser technology promises to be more efficient than current ones.

Researchers recently have made lasers using just a single photonic crystal microcavity in an attempt to outdo vertical cavity surface emitting lasers (VCSELs) [, which are way too weak.]

Vuckovic and Altug's laser successfully combines 81 microcavities (in a 9 by 9 array) to make it more powerful, and it operates with greater energy efficiency. In fact, [in Vuckovic and Altug experiments,] their laser was about 20 times as efficient as single-cavity lasers, putting out 100 times more power (12 millionths of a watt when pumped with 2.4 thousandths of a watt).

This research work has ben published by Optics Express under the title "Photonic crystal nanocavity array laser" (Vol. 13, No. 22, Pages 8819-8828, October 31, 2005). Here are two links to the abstract and to the full paper (PDF format, 10 pages, 1.19 MB).

Sources: Stanford University news release, November 7, 2005; and various web sites

You'll find related stories by following the links below.

• Networking
• Optics
• Photonics
• Physics
• Technology

For many years now, researchers have tried to make computers more user-friendly, sometimes with the use of avatars on screen trying to educate us. Results have never been really successful. But now, a researcher at Florida State University (FSU) is working to give computers a human 'face' and her computer-generated characters can be tailored to a specific audience. For example, she is using such agents to challenge young women's stereotypes about the engineering profession by employing non-stereotypical engineering 'mentors' like pseudo-women.

Here are some details about this FSU project.

"Up until now, the personal computer's potential to be a valuable teaching and learning tool has been stymied by its 'soulless' nature," said Amy L. Baylor, [an associate professor of instructional systems and director of FSU's Center for Research of Innovative Technologies for Learning (RITL).] "At RITL, we're using computers to simulate human beings in a controlled manner so we can investigate how they affect and persuade people."

Every day at RITL, researchers in the areas of instructional technology, human-computer interaction, communication, computer science and psychology work to develop innovative uses of technologies to support learning and performance.

And Baylor is achieving this goal with what she calls 'pedagogical agents.'

A pedagogical agent is an animated, three- dimensional character that serves as the "face" (and "interface") of the computer and that can mimic human emotional expressions, nonverbal communication and interactions.

Below is a animated version of one of these pedagogical agents named the 'Peer' model (Credit: FSU). Here is a link to an animated version of this agent. And you'll other examples on this demo site.

And these agents can be controlled and adapted to suit your needs in an optimal way.

"Unlike a human mentor, we can control all aspects of a pedagogical agent - its gender, age, ethnicity, personality, message, and interaction style - to represent the ideal persona for facilitating learning. This leads to all kinds of exciting possibilities for simulating and researching different teaching styles and instructional strategies."

And for more information about this project, you can read some of the publications of the researchers.

Sources: Florida State University news release, November 15, 2005; and various web sites

You'll find related stories by following the links below.

• AI
• Computers
• Education
• Future
• Human Computer Interface
• Psychology
• Vision and Visualization

With the purchase of Scientific-Atlanta, we've learned yesterday that Cisco Systems wants to come to our houses via our TVs. But Cisco also wants to go further and has already put a special router in orbit. The Cisco Low Earth Orbit (CLEO) project uses a customized version of Cisco's Mobile Access Router, typically used to connect computer equipment to an IP network. It was launched in 2003 and is still operational. With CLEO, military personnel can receive images from satellites on their laptops and send commands to the device over IP. Even if the router has been specifically modified for space, it's still using commercial off-the-shelf (COTS) technology which is far cheaper than the devices currently used by big satellite companies.

So far, Cisco has only said that CLEO was only a proof-of-concept exercise.

Routers in space hold promise for future satellite-based broadband technologies, which could make wide-area data network services ubiquitous and more robust than current satellite data services, says Rick Sanford, director of Cisco's Global Space Initiatives group. The use of commercial off-the-shelf (COTS) computing and network technology is also of interest to the government and aerospace industry.

Here is a brief history of the project.

Working with Surrey Satellite Technology, Cisco made CLEO available for launch in 2003 as piggyback cargo on the UK Disaster Monitoring Consortium satellite, part of a satellite network used to photograph hurricanes, wildfires and earthquakes from space.

Last year, CLEO was put to its big test, executed by the Air Force, Army and NASA's Glenn Research Centre at Vandenberg Air Force Base in California. In this test, military personnel sitting in a jeep used a laptop running special General Dynamics software to make IP-based contact with CLEO. From this Virtual Mission Operations Centre, laptop operators were able to download images from the satellite and send command-and-control signals to the device over IP.

Below is a diagram showing the CLEO network topology for the Vandenberg demonstration, which was a secure space-based network-centric operations network (Credit: NASA and Cisco).

Using satellites for communications is not new. So what Cisco is bringing to the party? Apparently, CLEO has two key advantages over current technologies used by satellite companies. (more…)