Association for Computing Machinery
Welcome to the October 12, 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


One Small Step for Robots, One Giant Leap for Robot-Kind?
CORDIS News (10/10/16)

The European Union-funded ROVINA project has developed a robot that can navigate and explore hard or dangerous places to reach, such as tunnels or mines, build textured 3D models--including a semantic interpretation--of the site, and return home on its own. The ROVINA group built a new robotic system to face challenges ranging from arduous terrain to scene interpretation barriers such as low lighting. The group integrated features such as digital preservation, autonomy and connectivity, 3D reconstruction and mapping, object detection and online learning, vision and perception, semantic analysis, and user interface design using a modular software design. Each module that performs a specific task interacts through a middleware interface. The group developed a prototype that was able to explore the catacombs of Rome and Naples. The researchers say the autonomous robot could be used by historians, archaeologists, construction engineers, and potentially virtual tourists. "With the software able to adjust to sensors for scanning larger areas, autonomous (or semi-autonomous) digitization becomes possible for [global-positioning system]-denied environments, such as complex factories floors or indoor environments," says ROVINA project coordinator Cyrill Stachniss.


How Virtual Reality Is Being Used to Deliver Mental Health Care
Australian Broadcasting Corporation (10/11/16) Michael Collett

Scientists around the world are starting to use virtual reality (VR) technology as part of mental health care. So far, the main use of VR has been in the area of exposure therapy for phobias and anxieties, according to Greg Wadley at Australia's University of Melbourne. For example, Wadley says commercially available VR headsets can put users in front of a virtual crowd, enabling them to confront their fear of public speaking in a controlled environment. Wadley also is working with a youth mental health clinic to develop VR treatments for young people suffering from psychosis and depression. The technology is designed to be used with the guidance of professionals, and Wadley says VR can help people overcome the two main limitations of in-person therapy--cost and accessibility. Meanwhile, University of Oxford researchers in the U.K. are helping people with paranoia deal with difficult social situations such as encountering strangers in crowded spaces. "Patients who fully tested out their fears in virtual reality were later much less distressed even when in a real-world situation, such as going to the local shop," the Oxford researchers note.


Here's How the White House Wants the U.S. to Approach AI R&D
TechCrunch (10/12/16) Devin Coldewey; Jonathan Shieber

The White House has issued two new reports that detail how the U.S. government wants to approach research and development (R&D) into artificial intelligence (AI) and what these initiatives should focus on. The reports cite such issues as the need for complete and unbiased data to ensure AI is fair, just, and accountable when making significant decisions about people. Another point is to cultivate interest in science, technology, engineering, and math among students through earlier education, possibly at primary or secondary school, while also teaching them ethics and pushing for more diversity. The government's strategic AI investment plan calls for federal emphasis on areas of societal importance that do not target consumer markets. The investments also should concentrate heavily on human-AI collaboration as opposed to the replacement of humans by AI. The plan's outline includes long-term AI research investments, developing human-AI collaboration methods, and addressing AI's ethical, legal, and societal ramifications. Other recommended steps include guaranteeing AI systems are safe and secure, developing shared public datasets and environments for AI training and testing, quantifying and assessing AI technologies via standards and benchmarks, and achieving a better comprehension of national AI R&D workforce needs.


Computer Scientist Publishes Manifesto for Expressive Algorithmic Music
Motherboard (10/09/16) Michael Byrne

Gerhard Widmer at Austria's Johannes Kepler University has published a manifesto for music information research (MIR) in the October issue of ACM Transactions on Intelligent Systems and Technology. He argues the MIR field, encompassing such technologies as music recommender systems and automated music recognition, should refocus so it can facilitate a "qualitative leap in musically intelligent systems." Widmer cites problems beyond computers' abilities--including differentiating between songs that a listener might find dull or interesting, or playing along with musicians in a musically sympathetic manner--as a starting point. The manifesto says computers must be imbued with capabilities that transcend processing music as data and patterns, with an emphasis on perception and appreciation. "This is now the time for the MIR community to embark on massive feature/representation learning endeavors--much like the current trend in image analysis, which starts to produce quite spectacular results," Widmer notes. "Given the computational and data-related demands, the MIR community should join forces and pool its resources, efforts, and learned models (in cases where the training data cannot be shared)--and indeed, it has already begun to do so." Widmer details a project, Con Expressione, designed to characterize and recognize music's expressive aspects via performance to develop new models for perception and generating expressive music.


Designing for 3D Printing
MIT News (10/11/16) Rachel Gordon

Researchers from the Massachusetts Institute of Technology's (MIT) Computer Science and Artificial Intelligence Laboratory this week will present "Foundry," a system for custom designing 3D printed objects using mixed materials. Traditional fabrication methods, even existing 3D printing, require objects of different materials to be manufactured separately and then assembled together. However, Foundry enables the variation of material properties at a fine resolution. "It's like Photoshop for 3D materials, allowing you to design objects made of new composite materials that have the optimal mechanical, thermal, and conductive properties that you need for a given task," says MIT postdoctoral researcher Kiril Vidimce. "You are only constrained by your creativity and your ideas on how to combine materials in novel ways." After designing an object in a traditional computer-aided design package, the designer creates an "operator graph" to finely determine composition. Vidimce wants Foundry to give rise to a community of designers who can share new composition-determining actions or "operators" to broaden the technology's production repertoire. The Foundry method will be presented as a paper this week at the ACM User Interface Software and Technology Symposium (UIST 2016) in Tokyo, Japan.


Looking for a Choice of Voices in AI Technology
The New York Times (10/09/16) Quentin Hardy

The field of conversational computing is starting to consider how the voices that computerized agents use reflect societal preconceptions about gender and race. The choice of the application's voice has a clear impact on design, branding, and human-machine interaction, and researchers such as University of Michigan professor and entrepreneur Jason Mars ultimately aim to make products that transcend cultural stereotypes. Google last week unveiled a slate of voice-based products using an app that speaks in tones similar to a young, educated woman "who understands cultural cues, and can wink at things," says Google's Ryan Germick. There are situations in which a male voice is the better option when the goal is to have people quickly determine they are conversing with a machine. However, artificial intelligence products are more often given a female voice. Justine Cassell, a professor at Carnegie Mellon University's Human-Computer Interaction Institute, says people need more than a computerized voice to ease interaction with machines. "We have to know that the [computer] is enough like us that it will run our program correctly," she says. Cassell designed an avatar with no definite gender or race for five-year-olds, and she says children of different backgrounds imprint their own gender and racial expectations on it.
View Full Article - May Require Free Registration | Return to Headlines | Share Facebook  LinkedIn  Twitter 


The White House Frontiers Conference
CCC Blog (10/11/16) Helen Wright

President Barack Obama on Thursday will host the White House Frontiers Conference, a national event co-hosted by the University of Pittsburgh and Carnegie Mellon University (CMU) to examine the future of innovation at home and abroad. The conference will focus on building U.S. capacity in science, technology, and innovation through several different tracks, including personal, local, national, global, and interplanetary. The Local track space focuses on smart inclusive communities, and Argonne National Laboratory researcher Charlie Catlett will give a talk titled "Instrumenting Cities: The array of Things and Open Data." In addition, University of Chicago researcher Rayid Ghani will speak about "Doing and Teaching Data Science for Social Good: Challenges, Opportunities, and Lessons Learned." The National track space is focused on artificial intelligence, machine learning, automation, and robotics. Some of the presenters on this track are Facebook's Yann LeCun, University of Illinois-Chicago researcher Tanya Berger-Wolf, CMU researcher Stephen F. Smith, and Johns Hopkins University researcher Suchi Saria. Finally, the Personal track space is focused on health innovation. Columbia University researcher Rafael Yuste will discuss the BRAIN Initiative, which brings together brain researchers and computer scientists for a scientific dialogue aimed at determining new opportunities for joint research.


First Demonstration of Brain-Inspired Device to Power Artificial Systems
University of Southampton (United Kingdom) (10/07/16)

University of Southampton researchers have demonstrated a nanoscale device, called a memristor, could be used to power artificial systems that can mimic the human brain. The team showed an artificial neural network (ANN) using memristor synapses can support sophisticated learning rules to carry out reversible learning of noisy input data. "Memristors offer a possible route towards that end by supporting many fundamental features of learning synapses [memory storage, online learning, computationally powerful learning rule implementation, two-terminal structure] in extremely compact volumes and at exceptionally low energy costs," says Southampton researcher Alex Serb. The researchers found the metal-oxide memristor array was capable of learning and re-learning input patterns in an unsupervised manner within a probabilistic winner-take-all network. The researchers note this technique is extremely useful for enabling low-power embedded processors that can process big data in real time without any prior knowledge of the data. "Our work establishes such a technological paradigm shift, proving that nanoscale memristors can indeed be used to formulate in-silico neural circuits for processing big data in real time; a key challenge of modern society," says Southampton researcher Themis Prodomakis. He says the new technology could be applied to a wide range of fields, including pervasive sensing technologies and real-time fuel monitoring in harsh or inaccessible environments.


How to Cut Cake Fairly and Finally Eat It Too
Quanta Magazine (10/06/16) Erica Klarreich

Researchers have devised an algorithm that can fairly divide a cake among any number of people, which serves as a metaphor for a wide spectrum of real-world challenges that involve dividing some continuous object among people who value its features differently. A previous algorithm generated an envy-free division, but its unbounded nature dictated it might have to run for any large number of steps, depending on the players' preferences. However, in April, Carnegie Mellon University postdoctoral researcher Simon Mackenzie and Haris Aziz at Australia's University of New South Wales posted a paper detailing a bounded envy-free algorithm that can cut a cake with a runtime based on the number of players only. Their algorithm repeatedly asks individual players to cut cake into "n" equal pieces, then asks other players to make trims and select pieces. However, it also executes additional steps, such as periodically exchanging portions of players' cake reserves in a carefully controlled manner for the purpose of increasing the number of domination relationships between players; this enables Mackenzie and Aziz to streamline the complexity of the problem. Although the algorithm's practical utility may be limited, the result "resets the subject" for experts who study cake-cutting, says Bryn Mawr College professor Walter Stromquist.


University to Unveil UK's First CAVE as Part of New Tech Hub
Prolific North (10/10/16) David Prior

Edge Hill University is preparing to unveil the cave automatic virtual environment (CAVE), the U.K.'s first super-immersive three-dimensional virtual environment, as part of its new Technology Hub. The CAVE is equipped with four screens, offering users the chance to fully immerse themselves in a virtual room. Users can input real data to replicate reality and experience real-life scenarios in 4K resolution. The Technology Hub aims to be a key resource for students, employers, and anyone else who wants to take advantage of this state-of-the-art technology. In addition, a Tech Sandpit will unite students and businesses to discuss challenges and ideas that can then be developed using expertise from across computer science and biotechnology. "Having this dedicated facility for technology shows the university's commitment to technological innovation, computing, and bioscience and reflects our success in producing employable, highly skilled graduates who contribute to the regional economy and beyond," says Edge Hill pro vice-chancellor Mark Allanson. He notes the Technology Hub also will be open to schools in areas where students are unlikely to enter higher education, or study science, technology, engineering, and math subjects.


Crypto Needs More Transparency, Researchers Warn
The Register (UK) (10/09/16) Richard Chirgwin

Researchers from the French Institute for Research in Computer Science and Automation and the University of Pennsylvania have called for cybersecurity officials to publish the seeds for the prime numbers on which their standards rely. The researchers demonstrated again that 1,024-bit primes can no longer be considered secure, by publishing an attack using "special number field sieve" mathematics to show an attacker could create a prime that looks secure but is not. "There are opaque, standardized 1,024-bit and 2,048-bit primes in wide use today that cannot be properly verified," the researchers warn. They say if any of the "hard-coded" primes were maliciously produced, it would be difficult to recognize by looking at the numbers, but factorization would be feasible. The researchers call for 2,048-bit keys to be based on "standardized primes" using published seeds, because too many cryptographic schemes do not provide any way to verify the seeds are not somehow backdoored. The researchers ran their experiments on France's Grid'5000 testbed, the University of Pennsylvania's Cisco UCS cluster, the University of Waterloo's CrySIP RIPPLE facility in Canada, and Eindhoven Technical University's Saber cluster in the Netherlands.


White House Slates $80M for City Tech Innovations
Computerworld (10/07/16) Matt Hamblen

The White House is expanding its Smart Cities Initiative with $80 million in new federal funds while also doubling the number of participating U.S. cities to more than 70. The investments will fund the development and deployment of emerging technologies to improve public safety, autonomous urban transportation, energy usage, and city services. U.S. National Science Foundation grants and investments comprise $60 million of the total funding, including $10 million to support access to gigabit-ready networks under the U.S. Ignite program. Meanwhile, a grant by the Smart Cities Council will help five cities conduct self-assessments and search for technological solutions. The winning cities will examine their current approaches with assistance from the TM Forum and develop new models based around driving citizen engagement. "Sensors in the Internet of Things and other technologies happening in society could form a powerful catalyst for change," says the TM Forum's Carl Piva. "Cities need to see what citizens want." The new funding is partly a response to a report from presidential advisers earlier this year recommending ways to maximize technology innovation, and the report emphasizes a need for a comprehensive information infrastructure and effective approaches to data integration and sharing.


Online Emancipation: Protecting Users From Algorithmic Bias
University of Nottingham (United Kingdom) (10/06/16) Lindsay Brooke

The U.K.'s Engineering and Physical Sciences Research Council is funding "UnBias: Emancipating Users Against Algorithmic Biases for a Trusted Digital Economy," a research project that will examine the user experience of algorithm-driven Internet services and the process of algorithm design. News feeds, search engines, and product recommendations increasingly use personalization algorithms, but whether Internet users know whether the information received is best for them is unknown. The UnBias project is an effort to establish a system of auditability and build trust and transparency. "Selections made by algorithms are commonly presented to consumers as if they are inherently free from human bias and fair, but there is no such thing as a neutral algorithm," says University of Nottingham research fellow Ansgar Koene. He notes the UnBias project will study the concerns and perspectives of Internet users with the aim of drawing up policy recommendations, ethical guidelines, and a "fairness toolkit." Koene says the research team will produce educational materials and resources to help young people better understand online environments. He notes the results of the two-year project will be disseminated through peer-reviewed journals as well as via community groups, schools, and youth organizations.


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


To submit feedback about ACM TechNews, contact: technews@hq.acm.org

Unsubscribe