Association for Computing Machinery
Welcome to the December 7, 2016 edition of ACM TechNews, providing timely information for IT professionals three times a week.

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


Infrastructure Data for Everyone
Technical University of Munich (Germany) (12/05/16)

Researchers at the Technical University of Munich (TUM) in Germany are collecting infrastructure information using an universally accessible open source platform called the OpenGridMap project. "What we are interested in is the electrical infrastructure: high-voltage and low-voltage power lines, transformer substations, wind turbines, and solar power plants," says OpenGridMap director Jose Rivera. The project involves a volunteer community roaming cities around the world, using the OpenGridMap application on their smartphones to share photos and locations with a TUM server to build a global map of electrical grids. "This is a prerequisite for the energy turnaround--not only here in Germany, but in all countries around the world," says TUM professor Hans-Arno Jacobsen. "You can only plan the restructuring of the energy supply if you know exactly where power lines are located and at which locations power from high-voltage lines is transformed and fed into the low-voltage networks." Rivera says the crowdsourcing app can help build a solid database despite the unaffordable cost of contracting a company to compile such a massive body of information. Among the potential uses for OpenGridMap Jacobson cites is improving a developing country's infrastructure.

When an Interface Becomes the Face
University of Vienna (Austria) (12/05/16) Glenda Hannibal

Social roboticist Glenda Hannibal at the University of Vienna in Austria considers how a future society supported by robots will transform human-digital interaction. "Could we imagine a future in which we do not use digital tools but rather interact with them as with independent beings?" she writes. Hannibal theorizes such a future would have interfaces functioning literally as faces. She traces social robotics' origins to the late 1990s with the advent of robot toys for children, and the move to develop socially functioning machines evolved as western countries saw technology as a solution to demographic shifts. "These robots are no longer seen as mere instruments, but rather as crucial interaction partners for those people in our society who might be in need of some special help or care," Hannibal notes. "But the ultimate goal in social robotics is a possibility of building more advanced robots that could eventually become true companions, and not only in the context of therapy or elder care." She says this ambition is reflected in the creation of robots as family companions. Hannibal predicts "humans will become more sensitive to how we can engage with technological interface (or face) and in turn robots will adapt to human needs."

Two-Day Conference: Shedding Light on Computing and Information Sciences Research
The Express Tribune (Pakistan) (12/07/16)

Nine universities from across Pakistan, including the Institute of Business Administration, the National University of Sciences and Technology, NED University, Mohammad Ali Jinnah University, and Bahria University from Karachi participated in a recent conference hosted by ACM and the PAF Karachi Institute of Economics and Technology (PAK-KIET) to encourage students to conduct research, publish their papers in international journals, and learn from seasoned academics on how to research. During the two-day conference, 16 research papers on different subjects were presented by the participants. "Research plays an important role in both academia and industry, which is why such activities help nurture a healthy research environment," says PAF-KIET president Tubrez Asif. A team of 10 organizing members hosted the international conference, the first of its kind, and they plan on conducting similar events every year. "Such conferences motivate students to research and polish their skills in the domain," says Muhammad Mustufa, one of the organizing members. He notes by presenting the research papers in front of researchers and fellows from other universities, students can learn to avoid the mistakes they would have never known about if they did not have this opportunity.

U.S. Lead in Quantum Computing 'Under Siege,' Says White House Cyber Adviser (12/06/16) Mohana Ravindranath

The U.S. needs to invest more in quantum computing, according to Tim Polk, assistant director of cybersecurity within the U.S. Office of Science and Technology Policy (OSTP). Although the federal government has been funding quantum research for decades, Polk says its global leadership position is currently "under siege," citing significant quantum programs in Canada, the U.K., the Netherlands, the European Union, and China. "Those programs are definitely going to pose a challenge to us and that's really important to us," Polk notes. "We believe [quantum] will be the foundation for so much economic development as well." Part of Polk's role at OSTP is to oversee the National Strategic Computing Initiative, an effort to promote research in high-performance computing, and a quantum research interagency working group. Polk says quantum algorithms eventually could help solve long-standing challenges in a wide range of fields. However, to really advance quantum, "we're going to need to be able to build small quantum devices that are off the shelf, commercially available, and that we can actually use not in the lab but actually in real life," he notes.

The Origin of Silicon Valley's Gender Problem
Quartz (12/06/16) Jenny Anderson

About 25 percent of boys and 24 percent of girls expect to be working in a science-related occupation in their 30s, according to the results of the Organization for Economic Cooperation and Development's (OECD) Program for International Student Assessment test. However, more boys expect to be engineers, scientists, or architects, while more girls hope to work in health, and less than 1 percent of girls expect to pursue a career in information technology (IT). Many in Silicon Valley believe an historic underrepresentation of women in IT is due to too few women studying the discipline compared to men. The OECD found a strong correlation between children's interest in pursuing a science career and their performance in the subject in school, with boys generally outperforming girls. The report also notes "gender stereotypes about scientists and about work in science-related occupations can discourage some students from engaging further with science." Schools are the starting point for reforming and dispelling these stereotypes and encouraging more gender parity in science. "Expanding students' awareness about the utility of science beyond teaching and research occupations can help build a more inclusive view of science, from which fewer students feel excluded," the OECD says.

17 Microsoft Researchers Offer Bold Computer Science Predictions for 2017 and 2027
GeekWire (12/05/16) Kurt Schlosser

Seventeen female Microsoft researchers on Monday offered their views on what they believe is in store for their fields next year and a decade later. Microsoft Research India researcher Kalika Bali expects speech and language technology applications to become increasingly multilingual in 2017. "We...will have systems that understand, process, and generate the language that an English-Spanish or a French-Arabic or a Hindi-English speaker uses when she effortlessly switches from one language to another, within the same conversation, chat, and sometimes even within the same sentence," she says. Bali also predicts that deep embedding language models within cognitive models will empower artificial intelligence with the ability to reason and converse with humans "relatively effortlessly." Meanwhile, Olya Ohrimenko, a researcher in Constructive Security and Programming Principles and Tools groups at MSR Cambridge in the U.K., says trusted hardware will build momentum for new apps and tools with robust security. "Only an encrypted form of our personal data will be used in medical and administrative analyses, machine-learning algorithms, and our daily online activities," Ohrimenko says. Microsoft Research's Oriana Riva, a researcher in the Hardware, Devices, and Experiences group, predicts more mobile computing systems will reconfigure themselves to support interactions free of graphical user interfaces, with an expanded relationship with the digital domain expected by 2027.

White House Boosts CSforAll Commitments for Computer Science Education Week
THE Journal (12/05/16) Sri Ravipati

In support of Computer Science Education Week 2016, the White House has announced new commitments involving implementing computer science (CS) and computational thinking (CT) in K-12 education that build on President Barack Obama's Computer Science for All (CSforAll) initiative. The largest new commitment comes from the U.S. National Science Foundation, which will invest $20 million in CSforAll: Researcher Practitioner Partnerships for fiscal year (FY) 2017, in addition to a $25-million investment in FY 2016. The program was created to better understand how to provide K-12 teachers with the preparation, professional development, and ongoing support they will need to teach and integrate CS and CT into their classrooms. Meanwhile, the U.S. Department of Education is expanding the scale and scope of its 21st Century Community Learning Centers, a $1.1-billion formula grant program that helps students in high-poverty and low-performing schools receive academic and enrichment opportunities in science, technology, engineering, and math. The program has expanded to 200 sites across 25 states and will involve several federal agencies. In addition, the U.S. National Science and Technology Council will develop a CSforAll strategic framework to guide federal efforts for the initiative. More than 250 different organizations have pledged support for CSforAll, including state and local government agencies, nonprofits, and private companies.

A Handful of Photos Yields a Mouthful of (Digital) Teeth
EurekAlert (12/05/16) Jennifer Liu

Disney researchers have developed a model-based method of realistically reconstructing teeth for digital actors and for medical applications. The new method relies on a few noninvasive photos or a short video of the mouth, and it can digitally reconstruct teeth even when some are obscured. "By combining creativity and innovation, this research continues Disney's rich legacy of leveraging technology to enhance the tools and systems used to create more realistic and believable digital actors for films or video games," says Disney Research vice president Markus Gross. The researchers developed their method by constructing a model of human teeth rows using high-resolution three-dimensional scans of 86 different teeth rows from the field of human dentistry. The team then used the data to develop a model of an average row of teeth, including natural variations in shape and spacing for each tooth. "Our algorithm only requires minimal user interaction and can operate on a set of individual, non-calibrated images, making teeth capture as easy and convenient as taking a few pictures or even a short video clip from a standard mobile phone," says Disney researcher Thabo Beeler. The researchers presented their method this week at the ACM SIGGRAPH Asia 2016 conference in Macao.

Safer, Less Vulnerable Software Is the Goal of New NIST Computer Publication
NIST News (12/05/16) Chad T. Boutin

The strategies compiled in a new U.S. National Institute of Standards and Technology (NIST) publication can lead to software with 100 times fewer vulnerabilities, according to computer scientists at the agency. They recommend coders adopt the approaches in the 60-page document, NIST Interagency Report (NISTIR) 8151: Dramatically Reducing Software Vulnerabilities. The report is an official response to a request for methods from the White House's Office of Science and Technology Policy. NIST's Paul E. Black and colleagues compiled the ideas while working with software assurance experts from private companies in the computer industry as well as several federal government agencies that generate a lot of code, including the Department of Defense and the National Aeronautics and Space Administration. The approaches include using math-based tools to verify the code will work properly, breaking up a computer's programs into modules so if one part fails, the program does not crash, connecting analysis tools for code that currently operate in isolation, using appropriate programming languages for the task that the code attempts to execute, and developing evolving and changing tactics for protecting code that is the target of cyberattacks.

New 'Printone' Tool Allows Users to Create 3D Printed Wind Instruments in Any Shape or Form
Dartmouth College (12/05/16) Amy D. Olson

Dartmouth College researchers have developed Printone, an interactive design tool enabling users to create functional 3D-printed wind instruments in any shape or form using interactive sound simulation feedback. The researchers used Printone to design 16 free-form wind instruments to play different melodies. The user initially inputs a 3D shape into the system, which then creates a hollow acoustic cavity. The user then chooses the area where they will blow into the instrument and selects the position and size of the finger holes. In addition, the scale of the object can be changed to create a specific range of notes. Users are guided by quick feedback from the simulation on how each change affects the notes produced. The Printone interface narrows the computational scope of the tool by modeling the instrument as a passive resonator in which the oscillation from the mouthpiece is excluded, and focuses on the dominant tone to predict frequency resonance, instead of considering the entire frequency spectrum. The research will be presented this week at the ACM SIGGRAPH Asia 2016 conference in Macao.

Computer Learns to Recognize Sounds by Watching Video
MIT News (12/02/16) Larry Hardesty

Computers already excel at recognizing speech and images, but they may be catching up in the area of natural sounds, such as crowds cheering or waves crashing. Researchers at the Massachusetts Institute of Technology's (MIT) Computer Science and Artificial Intelligence Laboratory have trained a sound-recognition system on video, instead of using hand-annotated data. During testing, the system was 13-percent to 15-percent more accurate in recognizing natural sounds than the previous best-performing system. On a dataset with 10 different sound categories the system could categorize sounds with 92-percent accuracy, and on a dataset with 50 categories it performed with 74-percent accuracy. On those same datasets, humans are 96-percent and 81-percent accurate, respectively. The researchers took advantage of the natural synchronization between vision and sound. "We scale up with tons of unlabeled video to learn to understand sound," says MIT graduate student Carl Vondrick. The team used existing computer-vision systems to categorize images in a video, and then their machine-learning system found correlations between those visual categories and natural sounds. The researchers say sound recognition could improve the context sensitivity of mobile devices and the situational awareness of autonomous robots.

Concerns as Face Recognition Tech Used to 'Identify' Criminals
New Scientist (12/01/16) Timothy Revell

Face-recognition technology can tell whether someone is a criminal or not, according to Xiaolin Wu at Canada's McMaster University in Canada and Xi Zhang at China's Shanghai Jiao Tong University. The researchers set out to disprove the idea there could be a link between someone's face and criminality. However, when shown a series of faces it has never encountered before, a system developed by Wu and Zhang picked out the ones belonging to convicted criminals. Exploiting machine learning, the team asked face-recognition software to guess whether a person in an ID-style picture was a criminal or not, and then fed it the answer. The researchers say it learned to tell the difference, and eventually achieved an accuracy of up to 90 percent. "We were very surprised by the result," Wu says. However, other face-recognition experts question Wu and Zhang's methodology, noting the criminal images came from a Chinese database of ID photos while the non-criminal images came from Internet profile pictures. Some experts say the technology should not be used in this manner in the real world.

A Wall-Crawling Roomba That Teaches Kids to Code
The Atlantic (11/30/16) Ed Yong

Harvard University researchers led by professor Radhika Nagpal have designed a robot that can climb over whiteboards to teach coding to people of all ages. The magnetically wheeled Root robot also uses sensors for light, color, and textures to navigate a maze, play music when it moves over different shapes, and follow trails left by another Root robot. The most basic level of Root's coding application starts with purely graphical languages to teach children using wordless icons to control the device, which gradually evolve to text-based languages for adults. "Many kits are limited to either very simple entry-level programming, or more complex but less accessible programming environments," says University of Southern California professor Maja Mataric. "Root seems to strike the balance by providing various levels of challenge, thereby being able to grow with the learner, and not be quickly outgrown." Nagpal says young children are skilled at understanding coding concepts, noting robots offer tangible feedback. "They write something and the robot moves; it's obvious when something has gone wrong," she says. Nagpal hopes the technology can help address a shortage of coders in the U.S.

Abstract News © Copyright 2016 INFORMATION, INC.
Powered by Information, Inc.

To submit feedback about ACM TechNews, contact: [email protected]