Welcome to the July 8, 2016 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
Makers of Self-Driving Cars Ask What to Do With Human Nature
The New York Times (07/07/16) John R. Quain
The push by some automakers to make fully autonomous vehicles a commercial reality is being tempered by others arguing against the complete elimination of human intervention. Google is one of several companies pursuing fully autonomous Level 4 cars needing zero driver input. However, there are still advocates for Level 3 cars, defined by the U.S. National Highway Traffic Safety Administration as those that can self-drive in specific circumstances, but still require a human motorist to be "available for occasional control, but with sufficiently comfortable transition time." Shane McLaughlin, director of the Virginia Polytechnic Institute and State University's Center for Automated Vehicle Systems, says Level 3 vehicles should not be so quickly dismissed. He says, "we can get the machine to give the person enough time to react." Such systems might feature in-vehicle video and infrared equipment monitoring the driver's attentiveness, while electronic horizon technologies also might extend drivers' reaction time by "seeing" farther down the road, with cars able to automatically talk to one another. Among the challenges of Level 4 cars cited by Level 3 proponents are making them capable of appropriately handling situations they have never encountered before. Level 3 supporters favor an incremental evolution of autonomous vehicles, as well as the development of industry standards for human-machine interaction in such cars.
Google Hopes to Thwart Quantum Computers From Cracking Today's Internet Encryption
IDG News Service (07/07/16) Michael Kan
Google is testing new cryptography methods that future computers might be unable to break. Google researcher Matt Braithwaite says the processing power offered by future quantum computers could be enough to "decrypt any Internet communication that was recorded today." Braithwaite notes that could affect the Transport Layer Security (TLS) protocol used when visiting websites, which means existing information meant to be secured for decades could suddenly become exposed. Google will deploy what it is calling "post-quantum cryptography" and will test it using the Chrome Canary browser. The researchers will test a cryptography algorithm called "New Hope" on a small number of connections between the browser and Google's computers, but the test will only last two years and afterwards they hope to replace the algorithm with something better. "The post-quantum algorithm might turn out to be breakable even with today's computer," Braithwaite says. "Alternatively, if the post-quantum algorithm turns out to be secure, then it'll protect the connection even against a future, quantum computer." Users of Chrome Canary can tell if the post-quantum algorithm is in use by checking the browser's security panel and looking for "CECPQ1" in the key exchange.
Building a Better Computer Bug Finder
NYU Tandon School of Engineering (07/06/2016)
Researchers at the New York University Tandon School of Engineering, the Massachusetts Institute of Technology Lincoln Laboratory, and Northeastern University say they have taken an unorthodox approach to measuring the efficacy of bug-finding tools. The team has developed a technique that intentionally adds vulnerabilities to a program's source code, and they believe the approach will ultimately help developers improve bug-finders. Large-Scale Automated Vulnerability Addition (LAVA) inserts known quantities of novel vulnerabilities that are synthetic yet possess many of the same attributes as computer bugs in the wild. The automated system makes judicious edits in real programs' source code. The result is hundreds of thousands of unstudied, highly realistic vulnerabilities that are inexpensive, span the execution lifetime of a program, are embedded in normal control and data flow, and manifest only for a small fraction of inputs lest they shut the entire program down. The team had to create many novel bugs in order to have a large enough body to study the strengths and shortcomings of bug-finding software. During tests, popular bug finders could detect only 2 percent of bugs created by LAVA.
A Celebration of Women in HPC
HPC Wire (07/07/16) Tiffany Trader
Recent events held by the Women in High-Performance Computing (WHPC) network focused on resolving the lack of gender diversity in HPC, with the organization estimating women constitute between 5 percent and 17 percent of HPC users, researchers, and conference attendees, and only about 25 percent of technology jobs. WHPC founder Toni Collis says the group seeks to widen the participation of women and other underrepresented groups in supercomputing conferences and to address bias and discrimination regarding women's HPC-related interests and capabilities. "We made an effort to target early career women, but universities send far fewer women to ISC (ISC High Performance, formerly the International Supercomputing Conference) than SC (the International Conference for High Performance Computing, Networking, Storage and Analysis)," Collis says. "I think they send their senior people to ISC and those are still primarily men. It's a problem that we need to change." Intel Exascale Lab director Marie-Christine Sawley says HPC has come increasingly closer to affecting people's daily lives over the last decade, a fact that makes broader gender representation vital. "The world is made of 51-percent females and the numbers show that in HPC and in technology, we are not close to even half of what the statistics would show," she says. Sawley notes areas in which HPC innovations are having significant impact include parcel delivery, pediatric cancer research, and improved genetic disease prenatal screening diagnostics.
Stanford Researchers Automate Process for Acquiring Detailed Building Information
Stanford News (06/29/16) Shara Tonn
Researchers at Stanford University have automated the process of obtaining detailed building information with a system dependent on existing three-dimensional (3D) sensing technologies. The sensors employ light to measure every feature of a building's interior to create a data file that captures the building's spatial geometry. The system is unique in how the Stanford process feeds the raw data file collected by the sensors into a new computer-vision algorithm. The algorithm automatically recognizes structural elements such as walls, columns, desks, and other furnishings. "This is the first time it's possible to do it at the scale of whole buildings, with hundreds of rooms," says Stanford professor Silvio Savarese. To train the computer-vision system, the researchers captured a massive volume of annotated 3D point cloud data that specified all kinds of building features. Via repetition, the system trained itself to identify different building elements. The end result is an algorithm that can analyze raw point cloud data from a complete building and, without human assistance, identify the rooms, enter each room, and detail the structural components and furniture. "This kind of geometric, contextual reasoning is one of the most innovative parts of the project," Savarese says.
Are Face Recognition Systems Accurate? Depends on Your Race
Technology Review (07/06/16) Mike Orcutt
The U.S. Government Accountability Office in June issued a report finding the Federal Bureau of Investigation (FBI) has not properly tested the accuracy of its face-matching systems or the massive network of state-level face matching databases it accesses. Studies show the facial-recognition systems used by the FBI and other police agencies have a built-in racial bias that is a result of how the systems are designed and the data on which they are trained. Although photos taken under controlled conditions with generally cooperative subjects can be nearly 95-percent accurate, images taken under less-than-ideal conditions can produce errors. The algorithms also can be biased based on the way they are trained, according to Michigan State University professor Anil Jain. Face matching software works by learning to recognize faces using training data. If a gender, age group, or race is underrepresented in the data, that will be reflected in the algorithm's performance. "If your training set is strongly biased toward a particular race, your algorithm will do better recognizing that race," says University of Texas at Dallas professor Alice O'Toole. In 2011, O'Toole conducted a study showing an algorithm developed in Western countries was better at recognizing Caucasian faces than it was at recognizing East Asian faces, while East Asian algorithms performed better on East Asian faces than on Caucasian faces.
A World of Help for Women Software Engineers
IEEE Spectrum (06/30/16) Tekla S. Perry
Women software engineers have started local organizations to support each all over the world, and U.S. Geological Survey senior software developer Veni Kunche wants to make it easier to find them. She has created a list of the major multi-chapter organizations, including Girls Develop It, Duchess Lesbians Who Tech, PyLadies, RailsBridge, Rails Girls, Ladies Learning Code, and Women Who Code. Kunche is using crowdsourcing to gather more information about less well-known communities, and has identified about 500 so far. In addition, Kunche has placed the organizations within an interactive map that enables users to zoom in on a community, browse the available groups, and click through to individual websites for more information. Kunche wants to make the interactive map more useful by enabling users to break out different categories of an organization and identify groups that focus on a specific programming language, or those that offer mentorships and help to entrepreneurs. She says the career path for a women in tech can be lonely. "Sometimes being in a room full of men, whether it's at work or a in a classroom, can make you feel like the odd one out, as though you don't belong there," Kunche says. "It is important for women to know that they are not alone."
Robot Eyes and Humans Fix on Different Things to Decode a Scene
New Scientist (06/29/16) Aviva Rutkin
Researchers at Facebook and the Virginia Polytechnic Institute and State University (Virginia Tech) are determining the differences between human minds and artificial intelligence-based (AI) machines through mapping human and AI visual attention. The attention maps are measurable in both humans and machines, and enable researchers to study how computers choose to decode a scene. Researchers asked human subjects to answer questions about a set of images that had been blurred, and the subjects then clicked around the screen to sharpen parts of the image. The clicks were logged to create a map of where a subject's attention was drawn, and the same test was conducted with neural-network machines that had been trained to interpret images. Although the neural networks were accurate in their answers, researchers found little overlap between machine and human attention. The results could be used by scientists looking to make their AI machines more closely resemble humans. "Machines do not seem to be looking at the same regions as humans, which suggests that we do not understand what they are basing their decisions on," says Virginia Tech researcher Dhruv Batra. "Can we make them more human-like, and will that translate to higher accuracy?"
EU Plans $2B Investment in Cybersecurity Research
IDG News Service (07/05/16) Peter Sayer
The European Union (EU) is contributing $500 million to fund research into cybersecurity, and is asking the private sector to contribute an additional $1.5 billion. The $2-billion cybersecurity public-private partnership (cPPP) aims to boost cross-border research into cybersecurity, and to aid the development of security products and services for a range of industries, according to the European Commission (EC). The EC already has taken steps toward this goal with the creation of the EU agency for network and information security (ENISA), the computer emergency response team, and Europol's European Cyber Crime Center. In addition, the EC is working on the Network and Information Security Directive, which will require EU member states to identify essential infrastructure operators and ensure they address the risk of cyberattack. The EC hopes these efforts will make the EU a world leader in cybersecurity, which could become a competitive advantage, as security is one of the fastest-growing information technology activities worldwide. The cPPP should be up and running by the third quarter of this year, and the initiative is currently accepting bids for its research funding, which is expected to start early next year.
No Need for Supercomputers
Lomonosov Moscow State University (06/28/16)
Researchers from Lomonosov Moscow State University's (MSU) Skobeltsyn Institute of Nuclear Physics have developed a method for calculating complex quantum-mechanical equations with personal computers, which they say is much faster than using supercomputers. The equations describe the scattering of a few quantum particles, representing a quantum-mechanical version of the Newtonian theory of three-body systems. The chief problem in solving such equations is calculating the integral kernel, which resembles a monitor screen with tens of billions of pixels, making their calculation via a good graphics processing unit possible. The researchers employed software created by Nvidia in conjunction with their own programs to divide their calculations over the many thousands of streams. MSU professor Vladimir Kukulin says the PC can perform in 15 minutes work that would usually take a supercomputer two to three days to execute. "We reached the speed we couldn't even dream of," he says. "The program computes 260 million of complex double integrals on a desktop computer within three seconds only." Kukulin says this breakthrough could be applicable to computing tasks in scientific areas such as plasma physics, electrodynamics, geophysics, and medicine.
DREAM Finish for Leukemia Challenge
Rice University (06/28/16) Mike Williams; David Ruth
The Dialogue for Reverse Engineering Assessment and Methods (DREAM) 9 challenge centered at Rice University revealed big data has a bright future in personalized medicine. As part of the international competition, 31 teams of computational researchers applied competing methods to a unique set of patient data gathered from hundreds of patients with acute myeloid leukemia. The challenge was to see how well the teams' algorithms could predict how patients responded to chemotherapy. The top-performing models were by Team EvoMed from Arizona State University and Team Chipmunks from the Ontario Institute for Cancer Research in Toronto. They were best able to predict patient response to therapy with an accuracy of close to 80 percent. Rice researcher Amina Qutub notes one interesting takeaway was that overall the 31 found it harder to predict outcomes for patients classified as "resistant to therapy" than for responsive patients. The median model prediction accuracy for resistant patients was 42 percent, versus 73 percent for responsive patients. The winning models were affected by the perturbation of signaling proteins known as phosphoinositide-3-kinase and NPM1. For Qutub's lab, the DREAM experience will provide a basis for experimentation on leukemia cell lines and testing whether specific sets of proteins offer a therapeutic advantage. The goal is to provide clinicians with a predictive tool to develop individualized treatment plans.
Microsoft Gets Hands-On With Gesture-Based Computer Interfaces
Gizmag (06/28/16) Michael Irving
Microsoft believes gesture tracking will be the next big thing in how humans interact with computing devices. A team at the company's Cambridge, U.K., lab is working to track all the possible configurations the human hand can form. The Handpose project is using the Kinect sensor packaged with an Xbox gaming console to track a user's hand movements in real time and display virtual versions that mimic everything real hands do. Meanwhile, Microsoft researchers in Redwood, WA, are experimenting with a system that is able to recognize that a physical button not connected to anything in the real world, has been pushed by reading the movement of the hand. Physical actual buttons and dials help make virtual interfaces feel more real, according to the researchers. Meanwhile, at Microsoft's Advanced Technologies Lab in Israel, researchers are working to incorporate hand gestures for various functions in their apps and programs so, for example, a computer could be locked by miming the turn of a key. The team says they fed millions of hand poses into a machine learning algorithm to train a system to recognize specific gestures, using hundreds of micro-artificial intelligence units to build a complete picture of a user's hand positions, as well as their intent. The algorithm scans the hands using a three-dimensional camera.
Energy-Efficient Security Mechanisms for Digital Currency
Ruhr University Bochum (Germany) (06/28/16) Julia Weiler
Bitcoin fraud prevention involves deploying security mechanisms that currently consume massive amounts of electricity, and Ruhr University Bochum professor Sebastian Faust suggests an alternative mechanism that uses less energy. His proposal involves a puzzle founded on storage space instead of on computing power, and solving it requires the user to initialize it in a central processing unit-intensive fashion. In the process, an immense percentage of hard disk storage space is employed, enabling a solution to be reached without any substantial computational expenditure, as long as sufficient storage space is available. Faust's system requires the user to sort a thread of digits in ascending order and save the sorted list; when he wants to publish the puzzle, he is requested to name the digit in a certain position in the list. If he saved the sorted list as instructed he can read the answer easily. A research team at the Massachusetts Institute of Technology and the Institute of Science and Technology Austria has broadened the proof-of-space concept and used it as the basis of a new digital currency. Faust's team also is exploring the security of smart contracts in the bitcoin network as another potential anti-fraud mechanism. In addition, he is developing a plan "to analyze the bitcoin system formally [using game theory concepts] and prove that it is secure."
Abstract News © Copyright 2016 INFORMATION, INC.
To submit feedback about ACM TechNews, contact: [email protected]
Current ACM Members: Unsubscribe/Change your email subscription by logging in at myACM.
Non-Members: Unsubscribe
|
