Association for Computing Machinery
Welcome to the June 12, 2015 edition of ACM TechNews, providing timely information for IT professionals three times a week.

Updated versions of the ACM TechNews mobile apps are available for Android phones and tablets (click here) and for iPhones (click here) and iPads (click here).

HEADLINES AT A GLANCE


Tipping Point in Transit
The New York Times (06/10/15) Farhad Manjoo

Many in academia and the transit industry believe the world is on the cusp of a tipping point in transit technology. M. Bart Herring, head of product management at Mercedes-Benz USA, says, "Cars and transportation will change more in the next 20 years than they've changed in the last 75 years." However, these changes may not come in the form many expect. When people think of the future of transportation, they envision automated vehicles, but most experts say such technology is at least a decade away, and even then could be too expensive for practical use on a wide scale. Experts say what is more likely in the near term is the proliferation of semi-autonomous technologies that can do some of the driving for people and help to dramatically reduce the risk of accidents. Changes in car technology will accompany the spread of services that enable people to get a car ride on demand or request private bus services. Meanwhile, smart transportation technologies spreading throughout the transportation infrastructure will help to make transit more efficient and timely. However, these changes might not all be positive, creating the potential for new kinds of accidents and providing benefit only to the wealthier classes.
View Full Article - May Require Free Registration | Return to Headlines | Share Facebook  LinkedIn  Twitter 


The Government's High-Tech Plan for Identifying You Based on Your Tattoos
The Washington Post (06/10/15) Andrea Peterson

One in five Americans has a tattoo, and on Monday, the U.S. National Institute of Standards and Technology (NIST) held a workshop reviewing the results of a challenge organized by the agency aimed at developing systems that could identify an individual based on their tattoos. The challenge was sponsored by the U.S. Federal Bureau of Investigation's Biometric Center of Excellence and brought together researchers from academia and the private sector. Several of the systems generated "hit rates well above 90 percent" in some of the simpler tests, such as basic tattoo detection and identification over time, according to NIST's Mei Ngan. However, other tests, such as matching sketches or computer graphics with a real tattoo, proved more challenging. Ngan says although such technologies could be used to help identify disaster victims, the primary use case for them is likely to be helping law enforcement track down criminals, who have tattoos at a higher rate than the general populace. Ngan says tattoos would likely not be used as a primary biometric like a fingerprint, since more than one person can have a tattoo, but as a way of narrowing down a potential list of suspects.
View Full Article - May Require Free Registration | Return to Headlines | Share Facebook  LinkedIn  Twitter 


Longstanding Problem Put to Rest
MIT News (06/10/15) Larry Hardesty

At the ACM Symposium on Theory of Computing (STOC) next week, Massachusetts Institute of Technology (MIT) researchers will present a paper showing the Wagner-Fischer algorithm is likely as efficient at measuring the difference between two genomes, texts, or anything that can be represented as a string of symbols as it can be. "There is a technical aspect of this paper, a certain gadget construction, that turns out to be very useful for other purposes as well," says MIT's Piotr Indyk. Wagner-Fischer works by assigning each symbol of one string to a column in a giant grid and each symbol of the other string to a row. The algorithm then fills in each square with the number of edits required to turn the string ending with the corresponding column into the string ending with the corresponding row. In computer terms, the algorithm runs in quadratic time. Although most researchers believe NP-complete problems take exponential time to solve, no one has been able to prove it. In their STOC paper, Indyk (who, along with Andrei Broder and Moses S. Charikar, received the 2012 ACM Paris Kanellakis Theory and Practice Award) and MIT student Arturs Backurs demonstrate that if it is possible to solve the edit-distance problem in less-than-quadratic time, then it is possible to solve an NP-complete problem in less-than-exponential time.


Coding Boot Camps Are on the Rise
Chronicle of Higher Education (06/10/15) Meg Bernhard

Coding boot camps could turn out as many as 16,000 graduates this year, up from 6,740 in 2014, according to a Course Report survey. The number of boot camps in the U.S. and Canada has risen to about 70, and the average cost of the courses is more than $11,000. Offering in-person instruction on programming and the opportunity to develop real-world programs, the boot camps are attracting more college graduates who want to improve their job prospects and people who are looking to change careers. Coding camps tend to have more of a vocational focus, compared to more theoretical and in-depth computer science courses, says analyst Trace Urdan. He notes college programs are unlikely to redesign their curricula in response to coding camps. Urdan expects the coding camp fad to level off in the years to come. Still, coding camps could potentially complement computer science departments' curricula if students enroll in coding programs first, notes Virginia Commonwealth University's Gardner Campbell.
View Full Article - May Require Paid Subscription | Return to Headlines | Share Facebook  LinkedIn  Twitter 


Traffic Hacking: Caution Light Is On
The New York Times (06/10/15) Nicole Perlroth

Cesar Cerrudo, a security researcher at IOActive Labs, has been trying for years to raise the alarm about the vulnerabilities affecting so-called "smart city" technologies. He has hacked smart traffic systems in cities across the world, including Washington, D.C., in an effort to raise awareness and put pressure on the vendors responsible for those systems. Now he is starting a nonprofit initiative, Security Smart Cities, with the goal of bringing together security researchers and public officials to tackle to vulnerabilities in increasingly connected cities. The market for smart city technologies is taking off, with Navigant Research estimating the market will reach $27.5 billion by 2023. Billions of new connected devices are expected to come online in the next few years, and researchers say the threats facing these devices are neither theoretical nor new. Almost a decade ago, in 2006, a pair of Los Angeles traffic engineers was accused of hacking the city's smart traffic light system. Although they only hit four intersections, their tampering caused gridlock that lasted four days. More recently, evidence emerged in 2013 that Russian hackers were targeting U.S. energy companies, possibly with an eye on disrupting smart energy grids. Yet, there is still no national legislation that would require municipalities or counties to address such threats, leaving numerous cities vulnerable.
View Full Article - May Require Paid Subscription | Return to Headlines | Share Facebook  LinkedIn  Twitter 


Machine Vision Algorithm Chooses the Most Creative Paintings in History
Technology Review (06/10/15)

In the art world, paintings deemed creative are believed to be novel and influential rather than derived from a previous influence.  To distinguish between creative and derivative paintings, Rutgers University researchers Ahmed Elgammal and Babak Saleh developed a program based on machine-vision technology that classifies images by the visual concepts they contain, which are called classemes. They can be low-level features such as color or texture, or higher-level features such as a walking figure.  The approach enables a machine-vision algorithm to analyze a picture and produce a list of classemes that describe it.  The algorithm developed by Elgammal and Saleh develops a list of up to 2,559 different classemes that is like a vector that defines a picture, which may be used to compare it against others analyzed in the same way.  The algorithm has access to two large art databases, one of which contains images and annotations on about 62,000 works of art. Elgammal and Saleh approached the history of art as a network in which each painting links to similar paintings in the past and future. Their technique enables them to make small changes to the algorithm to fine-tune definitions of creativity, which could impact the way art historians evaluate paintings.


Robot Eyes Will Benefit From Insect Vision
University of Adelaide (06/10/15)

A project in Australia is applying research in insect vision to an artificial intelligence system. Researchers at the University of Adelaide believe the visually guided behavior of flying insects could improve robotic visual systems. They note a dragonfly has low visual acuity and a tiny brain, but chases prey at speeds up to 60 km/h and successfully captures them at a rate of more than 97 percent. The team has developed an algorithm that helps emulate this visual tracking. "Detecting and tracking small objects against complex backgrounds is a highly challenging task," says Adelaide Ph.D. student Zahra Bagheri. "Instead of just trying to keep the target perfectly centered on its field of view, our system locks onto the background and lets the target move against it." The approach reduces distractions in the background, and gives time for the underlying brain-like motion processing to work. During testing, the system was as robust as the best target-tracking algorithms while running up to 20 times faster. The researchers now are transferring the algorithm to a bio-inspired, autonomous robot platform.


Stroke Patients Recover Arm Use With Virtual Reality
Pompeu Fabra University (Spain) (06/09/15)

Pompeu Fabra University (UPF) researchers have found using virtual reality to increase a stroke patient's confidence in using their paralyzed arm could facilitate the recovery process. Stroke patients with reduced muscle strength on one side of their body often under-use their affected limbs even though they still have some motor function. The researchers conducted a pilot study involving 20 hemiparetic stroke patients who worked with the Rehabilitation Gaming System, a program that enables users to control a virtual body via their own movements with the help of a Microsoft Kinect camera. Participants were asked to reach targets appearing in a virtual environment over several blocks of trials. In some of the trials, the researchers enhanced the movement of the paretic limb's virtual representation. The researchers found after seeing the enhanced movement, the patients started using their paretic limbs more frequently, suggesting "that changing patients' beliefs on their capabilities significantly improves the use of their paretic limb," says UPF researcher Belen Rubio. Following the intervention, the researchers found there was a significantly higher probability the patient would select their paretic limb to reach toward a virtual target, even if there was no amplification of movement in that session.


Strong Constraint Exists on One-Way Street That Delivers Optical Signals to Computers
Stanford Report (06/08/15) James Urton

Stanford University researchers led by professor Shanhui Fan and researcher Yu Shi have found a strong theoretical limitation to the miniaturization of optical technology, a breakthrough that will guide researchers searching for ways to build light-based computer chips. The researchers studied a device called a nonlinear isolator, with the hope that it would allow information from beams of light to travel in only the forward direction, while prohibiting transmission in the backward direction. The researchers note this type of device would simplify data transmission on computer chips. However, the researchers found backward leakage that had previously gone unrecognized in this class of device, and backed up their conclusions with computer simulations. They found in nonlinear isolators, when the forward beam is on, the backward beam leaks through the device. The research shows engineers will need to try to develop new devices that can keep optical information flowing in one direction, but not the other. The researchers hope their discovery will renew interest in other methods to create one-way streets for optical communication on a computer chip.


NASA Tests DARPA Challenge Robot for Space Manufacturing
Computerworld (06/09/15) Sharon Gaudin

The U.S. National Aeronautics and Space Administration's (NASA) Jet Propulsion Lab (JPL) has big plans for Robosimian, a four-legged robot that took fifth place in last weekend's finals of the U.S. Defense Advanced Research Projects Agency's Robotics Challenge. The space agency envisions the robot building parts for the International Space Station and satellites in space. Researchers say its SpiderFab project has the potential to change the way spacecraft are built and deployed. Robosimian can stand upright on two legs and use its other two limbs as arms with hands capable of grasping a lever, holding a tool, or turning a valve. "The dexterity and mobility capabilities that come along with Robosimian could be adapted for zero-g [gravity] environments," says JPL researcher Brett Kennedy. He notes Robosimian is expected to look much like it does now after a redesign, and a test in space is likely a few years away. Assembly work in space would reduce costs for NASA. "The most immediate research we're going to be doing...would be more along the lines of building very large telescopes or fuel depots for satellites," Kennedy says. "Currently, the largest telescopes we can build are limited by the size of the rocket that launches them."


Crossing a Critical Threshold
Lehigh University (06/05/15) Kurt Pfitzer

Researchers from Lehigh University, Japan, and Canada say they have made progress toward the goal of all-optical data transmission through a single crystal waveguide in glass. The researchers said they used ultrafast femtosecond lasers to produce a three-dimensional (3D) single crystal capable of guiding light waves through glass with little loss of light. They say this finding will boost ongoing efforts to develop photonic integrated circuits (PICs) that are smaller, less expensive, more energy-efficient, and more reliable than existing networks, which are based on discrete optoelectronic components such as waveguides, splitters, modulators, filters, and amplifiers to transport optical signals. To fabricate 3D PICs, the researchers say it is necessary to first prevent light from scattering as it is being transmitted and then to transmit and manipulate light signals fast enough to handle increasingly large quantities of data. To pattern crystals in glass, the researchers employed femtosecond lasers, whose high-intensity laser pulse enables nonlinear optical absorption, giving scientists control over where the laser is focused and where light is absorbed. "The ability to pattern nonlinear optical crystals in glass is therefore essential for 3D laser-fabrication of PICs to achieve its full potential," the researchers say.


Intelligent Machines Part 1: Big Data, Machine Learning and the Future
CIO Australia (06/04/15) Rebecca Merret

Artificial intelligence (AI) technologies, especially the latest innovations in machine learning, are being used by major players across the tech industry, defense sector, and others to tackle a broad array of challenges. At Facebook, AI technology underlies many of the social network's most basic features, such as automatically tagging people in photos and identifying content that could be relevant to each user's interests. The company also is working to make its AI technology smarter; this year it announced it was able to train an AI network to read a short version of "Lord of the Rings" and answer questions about the narrative. Meanwhile, Baidu is working to improve the AI technology underlying its search engine's speech-recognition capabilities, with the goal of enabling users to easily "talk" to their smartphones and other devices. Australia's Defense Science and Technology Organization is using AI to control autonomous drones operating on land, sea, and air, and to process intelligence materials such as satellite-based mapping data. National ICT Australia is using computer-vision AI to improve the quality of a "bionic eye" it is developing to help people suffering from macular degeneration. Finally, IBM is teaching its Watson supercomputer to ask questions of human experts to help it learn how to carry out tasks such as sorting drug discovery documents.


Abstract News © Copyright 2015 INFORMATION, INC.
Powered by Information, Inc.


To submit feedback about ACM TechNews, contact: technews@hq.acm.org
Current ACM Members: Unsubscribe/Change your email subscription by logging in at myACM.
Non-Members: Unsubscribe