Welcome to the June 9, 2021 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."
|
|
|
Leader in Power-Efficient Computer Architecture Receives Eckert-Mauchly Award
ACM
June 7, 2021
ACM and the IEEE Computer Society have named Princeton University's Margaret Martonosi recipient of the 2021 Eckert-Mauchly Award for her role in designing, modeling, and confirming power-efficient computer architecture. Martonosi was a pioneer in the design and modeling of power-aware microarchitectures, including the use of narrow bit-widths, thermal-issue modeling and response, and conducting power estimation. Martonosi and co-author David Brooks greatly reduced processor power consumption through a paper that introduced two optimizations. She and Brooks later demonstrated that a central processing unit can be engineered for a much lower maximum power rating, and minimally impact typical applications. Martonosi also presented the potential of fast, early-stage, formal methods to verify the correctness of memory consistency model deployment, through the Check verification tool suite.
|
FBI Secretly Ran Anom Messaging Platform, Yielding Hundreds of Arrests in Global Sting
The Wall Street Journal
Byron Tau; James Marson
June 8, 2021
Global authorities have arrested hundreds of suspected members of international criminal networks by tricking them into using Anom, an encrypted communications platform run by the U.S. Federal Bureau of Investigation (FBI). A bureau-led international law enforcement coalition monitored Anom, which makes and distributes mobile phones equipped with a covert communications application service. The FBI's San Diego field office co-opted Anom in 2018; with the cooperation of a confidential source, the FBI and its law-enforcement partners secretly embedded the ability to covertly intercept and decrypt messages. FBI special agent Suzanne Turner said, “The immense and unprecedented success of Operation Trojan Shield should be a warning to international criminal organizations—your criminal communications may not be secure; and you can count on law enforcement world-wide working together to combat dangerous crime that crosses international borders.”
*May Require Paid Registration
|
U.S. Senate Passes Bill to Encourage Tech Competition, Especially with China
NPR
June 8, 2021
The U.S. Senate passed legislation to ramp up semiconductor production and development of advanced technology amid intensifying global competition, especially from China. The bill allocates $50 billion to the Commerce Department to prop up chip development and fabrication via research and incentive programs previously greenlit by Congress. One provision would establish a new artificial intelligence- and quantum science-focused directorate within the National Science Foundation, authorizing up to $29 billion over five years for the new unit, and another $52 billion for its initiatives. Said Senate Majority Leader Chuck Schumer (D-NY), "Whoever wins the race to the technologies of the future is going to be the global economic leader, with profound consequences for foreign policy and national security as well." The House Science Committee is expected to consider that chamber's version of the legislation soon.
|
Keeping a Closer Eye on Seabirds with Drones
Duke University Nicholas School of the Environment
June 8, 2021
Using drones and artificial intelligence to monitor seabird colonies is less expensive, labor-intensive, and error-prone than on-the-ground surveillance, according to a study by scientists at Duke University, the Wildlife Conservation Society (WCS), North Carolina State University, and Falklands Conservation. The researchers analyzed over 10,000 drone photos of mixed seabird colonies in the Malvinas/Falkland Islands off Argentina's coast using a deep learning algorithm, which identified and counted albatrosses in the colonies with 97% accuracy, and penguins with 87% accuracy. In all, the researchers found the automated counts were within 5% of human counts about 90% of the time. Said Duke’s Madeline C. Hayes said, “Using drone surveys and deep learning gives us an alternative that is remarkably accurate, less disruptive, and significantly easier.”
|
Less Nosy Smart Speakers
The Michigan Engineer News Center
Gabe Cherry
June 8, 2021
University of Michigan (U-M) researchers have designed a device called PrivacyMic to reduce eavesdropping by smart speakers by notifying household devices of important data without recording speech. PrivacyMic pieces together ambient ultrasonic information that indicate when its services are required. The system compresses these ultrasonic signatures into smaller files that feature key bits of information, while removing noise within the range of human hearing. The researchers showed that PrivacyMic was more than 95% accurate in identifying household and office activities. U-M’s Alanson Sample said, “What we’ve found is that you can have a system that understands what’s going on, and a hard guarantee that it will never record any audible information.”
|
Early Endeavors on the Path to Reliable Quantum Machine Learning
ETH Zurich (Switzerland)
June 8, 2021
An international research team led by computer scientists at ETH Zurich in Switzerland used the theory of quantum hypothesis testing to develop a new approach that could pave the way for reliable quantum-based machine learning models. The approach can identify the threshold above which the assignments of a quantum classification algorithm are guaranteed to be reliable and robust. ETH's Zhikuan Zhao said, "When we realized that quantum algorithms, like classical algorithms, are prone to errors and perturbations, we asked ourselves how we can estimate these sources of errors and perturbations for certain machine learning tasks, and how we can guarantee the robustness and reliability of the chosen method. If we know this, we can trust the computational results, even if they are noisy." The researchers found their method can determine the error tolerance of a computation for both hacking attacks and natural errors.
|
Feds Recover More Than $2 Million in Ransomware Payments from Colonial Pipeline Hackers
The Washington Post
Ellen Nakashima
June 7, 2021
U.S. officials say more than $2 million in cryptocurrency payments to the hackers who held Colonial Pipeline hostage in May has been recovered, marking the first recovery by the U.S. Department of Justice's new ransomware task force. Federal Bureau of Investigation deputy director Paul Abbate said the bureau seized proceeds paid to the Russian DarkSide hacker ring from a digital "wallet" containing the ransom, after securing a warrant from a federal judge. An affidavit said the bureau acquired the wallet's "private key," while officials have not disclosed how it was obtained. In announcing the seizure, Deputy Attorney General Lisa Monaco said, "The sophisticated use of technology to hold businesses and even whole cities hostage for profit is decidedly a 21st century challenge. But the adage, ‘follow the money’, still applies."
|
Super Productive 3D Bioprinter Could Speed Drug Development
UC San Diego Jacobs School of Engineering
June 8, 2021
A three-dimensional (3D) bioprinter developed by researchers at the University of California San Diego (UCSD) that can produce large batches of custom biological tissues at record speed could accelerate drug development. The new bioprinting method can produce a tissue sample in just 10 seconds, compared to hours with traditional methods. The researchers designed 3D models of biological structures on a computer, which slices the models into 2D snapshots and transfers them to millions of microscopic-sized mirrors, which are digitally controlled to project patterns of violet light in the form of these snapshots. After the light patterns are shined onto a solution that solidifies upon exposure to light, the structure is printed a layer at a time in a continuous fashion. UCSD’s Shangting You said, “What we are developing here are complex 3D cell culture systems that will more closely mimic actual human tissues, and that can hopefully improve the success rate of drug development.”
|
Scientists Can Predict How Well a Stroke Survivor Will Recover Language Skills Using Computer Simulations of the Brain
The Brink (Boston University)
Jessica Colarossi
June 3, 2021
Scientists at Boston University and the University of Texas at Austin used computer simulations of the brain to predict the extent to which bilingual Hispanic stroke survivors will recover language skills. The neural network models simulate the brain of a bilingual person with language impairment, and their brain's response to therapy in English and Spanish. The model can highlight the optimal language to focus on in rehabilitation, as well as forecasting post-therapy results. The models were able to predict therapeutic effects accurately in the treated language, which implied that the simulations could inform healthcare providers' rehabilitation plans.
|
Want Your Nails Done? Let a Robot Do It.
The New York Times
Ellen Rosen
June 1, 2021
Three startups are developing robotics for automatically manicuring fingernails using technology that combines fingernail-painting hardware with machine learning software to differentiate the nail from the surrounding skin. The companies—Nimble, Clockwork, and Coral—use a database of thousands of nail shapes recorded by cameras when customers have manicures. Clockwork is testing a tabletop device at a pop-up location in San Francisco; the device incorporates computer vision and artificial intelligence (AI), while a gantry applies polish via multiaxis movements. Clockwork co-founder Aaron Feldstein said the product has a plastic-tipped cartridge to prevent piercing fingers, and thwarts hacking by not being connected to the Internet. Meanwhile, Nimble founder Omri Moran said his startup's technology blends computer vision, AI, and a robotic arm to polish and dry nails within 10 minutes, in a device about the size of a toaster.
*May Require Paid Registration
|
Researchers Fine-Tune Control Over AI Image Generation
NC State University News
Matt Shipman
June 1, 2021
Refined control over artificial intelligence (AI)-driven conditional image generation by North Carolina State University (NC State) researchers has potential for use in fields ranging from autonomous robotics to AI training. NC State's Tianfu Wu said, "Like previous approaches, ours allows users to have the system generate an image based on a specific set of conditions. But ours also allows you to retain that image and add to it." The approach also can rig specific components to be identifiably the same, but shifted position or somehow altered. In testing the approach using the COCO-Stuff and the Visual Genome datasets, the technique bested previous state-of-the-art image generation methods. Wu suggested applications for the technique like helping autonomous robots "imagine" the appearance of an end result before undertaking a given task, or producing images for AI training.
|
SeKVM Makes Cloud Computing Provably Secure
IEEE Spectrum
Charles Q. Choi
June 7, 2021
Columbia University researchers say they have designed the first hypervisor that can ensure secure cloud computing. SeKVM is a secure version of the popular KVM open source hypervisor that cuts the verification workload through what the researchers call microverification. Microverification reduces a hypervisor to a small core and an array of untrusted services, then demonstrates security by verifying the core exclusively; the core lacks exploitable vulnerabilities, and mediates the hypervisor's interactions with virtual machines (VMs) so one compromised VM cannot impact others. The researchers developed microverification-based MicroV software to authenticate large commercial-grade multi-processor hypervisors. Columbia's Ronghui Gu said, "SeKVM will lay a foundation for future innovations in systems verification and lead to a new generation of cyber-resilient systems software."
|
|
