Welcome to the July 19, 2019 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).
To view "Headlines At A Glance," hit the link labeled "Click here to view this online" found at the top of the page in the html version.
The online version now has a button at the top labeled "Show Headlines."
|
|
Facial Recognition Tech Is Growing Stronger, Thanks to Your Face
The New York Times Cade Metz July 13, 2019
Many databases of facial images are compiled by companies and researchers without the knowledge of the owners of those faces. Privacy advocates have observed that repositories of facial images assembled by Microsoft, Stanford University, and others hold millions of images. Facebook and Google, likely to have amassed the largest facial datasets, purportedly do not distribute their image troves, but academics, activists, and public papers demonstrate that other organizations have shared theirs with researchers, governments, and private enterprises in Australia, China, India, Singapore, and Switzerland for training artificial intelligence. However, there are concerns that technologies enabled by these datasets are being used in potentially invasive ways, such as when recently released documents showed Immigration and Customs Enforcement officials scanned motorists’ photos with facial recognition technology to identify undocumented immigrants.
|
Cyberattacks Inflict Deep Harm at Technology-Rich Schools
Associated Press Michael Melia July 16, 2019
Schools that have adopted education technology increasingly are falling prey to the types of cyberattacks once reserved for banks and other institutions holding sensitive data. The resulting disruptions have pushed some districts to halt the use of smart boards, student laptops, and other Internet-connected tools. The Federal Bureau of Investigation (FBI) says schools are being targeted for the sensitive data they hold, and because they provide critical public services. North Dakota chief information security officer Sean Wiese said school networks “may be considered easy targets because they’re a little bit more open than your traditional corporate culture. I do feel that is changing, just not quickly enough.”
|
Automated System Generates Robotic Parts for Novel Tasks
MIT News Rob Matheson July 12, 2019
Researchers at the Massachusetts Institute of Technology (MIT) have developed an automated system that can design and print in three dimensions complex robotic parts optimized according to a large number of specifications—a task that would be staggeringly difficult for humans to do manually. The researchers demonstrated the system by fabricating actuators that show different monochrome images at different angles. For example, one actuator portrays a Vincent Van Gogh portrait when laid flat, and "The Scream" by Edvard Munch when it is tilted at an angle and activated. The team also three-dimensionally (3D) printed water lilies with petals equipped with arrays of actuators and hinges, that folds up in response to magnetic fields run through conductive fluids. Said MIT’s Subramanian Sundaram, "Our ultimate goal is to automatically find an optimal design for any problem, and then use the output of our optimized design to fabricate it."
|
Peugeot-Maker PSA Testing Autonomous Driving in Spain
CNBC Anmar Frangoul July 17, 2019
Groupe PSA, the auto giant behind the Peugeot, Opel, and Citroen brands, has joined forces with the Automotive Technology Centre of Galicia (CTAG) to "advance the development of autonomous driving" with tests focusing on vehicle-to-infrastructure-communications. The tests in Vigo, Spain, aim to determine how vehicles can communicate with "surrounding infrastructure in a complex urban environment," with a focus on such areas as automated valet parking, autonomous driving in urban areas, and "optimal speed regulation" when approaching traffic lights. The tests are part of the European AUTOPILOT (Automated Driving Progressed by Internet of Things) project, which seeks to improve automated driving using Internet of Things technologies. Said Groupe PSA’s Ignacio Bueno, "These initiatives bring together the various components of the ecosystem that need to be created in order to deploy the technologies enabling connected, autonomous vehicles."
|
Microsoft Eyes Mozilla's Rust to Obliterate C++ Memory Security Flaws
CSO Australia Liam Tung July 18, 2019
Microsoft is encouraging developers to adopt Mozilla's "safe" Rust programming language, in an effort to address the memory corruption bugs in C and C++ code that are the main source of modern software vulnerabilities. In a blog post, researchers at the Microsoft Security Response Center (MSRC) outlined the case for developers using "memory-safe languages," highlighting Rust as a solution to help developers focus on feature development, rather than dealing with bugs they introduce while coding in C and C++. MSRC’s Gavin Thomas called Rust "one of the most promising newer systems programming languages," since it offers both the speed of C++ and the safety of Microsoft's .NET C# (C sharp) language.
|
The AI Technique that Could Imbue Machines with the Ability to Reason
Technology Review Karen Hao July 12, 2019
While deep learning algorithms have made great strides in giving machines perceptual abilities like vision, they have fallen short of giving them human-like reasoning skills. Yann LeCun, chief AI scientist at Facebook, a professor at New York University and 2018 ACM A.M. Turing Award co-recipient, suggested during a recent ACM webinar that the deep learning subcategory known as unsupervised learning could help the technology overcome this hurdle. LeCun thinks researchers should focus on temporal prediction by training large neural networks to predict the second half of a video when given the first. Ultimately, LeCun said, unsupervised learning will help computers develop a model of the world that can be used to predict future states of the world.
|
Drones Could Herd Rhinos From Poaching Hotspots
New Scientist Ruby Prosser Scully July 17, 2019
Researchers at the University of Brighton (UoB) in the U.K. have found that drones and harsh siren sounds were most effective in getting endangered southern white rhinos to leave poaching hotspots in national parks and reserves in South Africa. The UoB study considered alternatives to the vehicle and foot patrols typically used by these parks and reserves to keep the animals away from poachers. Said UoB's Samuel Penny, "We found the drones to be a much more effective tool at manipulating rhino movement than the siren, mainly due to their long transmission range and capability of pursuing rhinos over long distances. The drone's maneuverability also allows for more precise control of rhino movement, as it can be flown into positions in response to changes in rhino direction."
|
Researchers Build Transistor-Like Gate for Quantum Information Processing – with Qudits
Purdue University News Kayla Wiles July 16, 2019
Purdue University researchers used qudits to develop what could be a quantum version of a transistor, known as a gate. Qubits exist only in superpositions of 0 and 1 states, while qudits exist in multiple states, such as 0 and 1 and 2, meaning that more data can be encoded and processed. The new gate would be inherently more efficient than qubit gates, but it would also be more stable because the researchers packed the qudits into photons, which are not easily disturbed. The gate creates one of the largest entangled states of quantum particles ever achieved. Said Purdue researcher Andrew Weiner, "This gate allows us to manipulate information in a predictable and deterministic way, which means that it could perform the operations necessary for certain quantum information processing tasks."
|
Researchers Launch First Knowledge Graph on European Public Procurement
University of Southampton July 15, 2019
A newly released knowledge graph was designed by open data specialists to boost procurement data analytics and decision-making capabilities across Europe. Researchers working with the "TheyBuyForYou" project integrated tender and company data to develop the open source knowledge graph for public procurement. Said University of Southampton's Elena Simperl, "Knowledge graphs bring together data from a variety of sources into a common format that can be easily extended and reused by organizations. By releasing the graph open source, we hope to encourage developers to use it in their own products and give us feedback on how we could improve it."
|
Symantec Reveals WhatsApp and Telegram Exploit that Gives Hackers Access to Your Personal Media
VentureBeat Dean Takahashi July 15, 2019
Symantec has identified an exploit that could allow WhatsApp and Telegram media files to be exposed and manipulated by malicious actors. The Media File Jacking security flaw is a product of the time difference between when media files received through the apps are written to a disk, and when they are loaded in an app's chat user interface. Symantec's malware detection engines can detect apps that exploit this vulnerability, which a malicious actor could take advantage of to misuse and manipulate sensitive information. Symantec said the exploit is especially concerning because many people believe security mechanisms like end-to-end encryption protect the latest IM apps from manipulation and privacy issues.
|
Researchers Improve AI that Can Tell from Your Voice if You’re Depressed
Folio (University of Alberta) Katie Willis July 11, 2019
Researchers at the University of Alberta in Canada have improved an artificial intelligence (AI) system to detect whether a person is depressed by analyzing the sound of their voice. The researchers developed a method combining several machine learning algorithms to recognize depression more accurately from acoustic cues. The tool was trained on two standard benchmark sets of audio recordings ranging from five to 50 minutes long. In addition, the team built on past studies suggesting the timbre of a person's voice contains information about his or her mood. The ultimate goal is to develop useful applications from the technology, such as helping people reflect on their moods over time, or working with mental healthcare providers.
|
|