Welcome to the September 27, 2023, 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."
|
|
|
Inflation Bites U.S. Engineering Salaries
IEEE Spectrum
Tekla S. Perry
September 26, 2023
The IEEE-USA 2023 Salary and Benefits Survey indicates U.S. engineering salary growth lagged inflation for the second consecutive year, marking the first multiyear income decline since the 2000s. The median income of 3,992 polled U.S. engineers and other technology professionals who were IEEE members increased 6% from $160,097 to $169,000 between 2021 and 2022. However, average salaries in real dollars fell $3,585 in 2022 following a $3,723 dip in 2021. Although the gender gap contracted by $7,100 last year, men still earn an average $26,800 per year more than women.
|
Qubit Circuit Enables Quantum Operations with Higher Accuracy
MIT News
Adam Zewe
September 25, 2023
Massachusetts Institute of Technology (MIT) researchers significantly upgraded the performance of operations between quantum bits (qubits) via the new fluxonium superconducting qubit framework. The fluxonium-transmon-fluxonium architecture joins two fluxonium qubits with a tunable transmon coupler to perform highly accurate gate operations while restraining error-introducing background interaction. The fluxonium qubits support millisecond-exceeding coherence times, topping those of traditional transmon qubits roughly 10-fold. The researchers demonstrated two-qubit gates with 99.9% fidelity and single-qubit gates with 99.99% accuracy, and used extensible fabrication to deploy this circuit on a chip. MIT's Bharath Kannan said, "We strongly believe that this architecture, or something like this using fluxonium qubits, shows great promise in terms of actually building a useful, fault-tolerant quantum computer."
|
AI, Robots, Satellite Sensors Helping in the Fight Against Wildfires
Bloomberg
Coco Liu
September 25, 2023
Artificial intelligence (AI) and other technological tools increasingly are being used to identify and fight wildfires. Pano AI uses panoramic cameras to snap minute-by-minute photos and leverages an AI algorithm to analyze those images to detect signs of a fire. Meanwhile, OroraTech uses infrared imaging sensors on more than 20 satellites to spot fires across the globe from space with its ability to see through forest canopy and thick smoke. French startup Shark Robotics builds robots that human firefighters can operate from up to one kilometer (0.6 miles) away. They can watch the flames via the robot’s thermal camera and use water to battle the blaze.
*May Require Paid Registration
|
Face Orientation Estimation Using Depth-Gyro Sensor
Shibaura Institute of Technology (Japan)
September 22, 2023
Researchers at Japan's Shibaura Institute of Technology (SIT) used a depth-gyro sensor to collect point cloud data for various face orientations in order to train a deep learning-based classification model in face orientation estimation. The researchers altered a face's horizontal angle relative to the camera from more than 90 degrees to less than 90 degrees using step sizes of 30, 22.5, 18, and 15 degrees between them, supporting more than seven representative classes of face direction. Said SIT's Chinthaka Premachandra, "Precise training data for each orientation was obtained from the integration of the depth and gyro sensors, which reduce the number of point cloud samples required for constructing the classification model. Furthermore, applying a weight reduction process to reduce the weight of point cloud data enhanced training efficiency and resulted in fast face orientation estimation."
|
Scientists Closer to Finding a Test for Long COVID
Gizmodo
Ed Cara
September 26, 2023
A multi-institutional team of scientists thinks it may have discovered biomarkers of long COVID that could lay the groundwork for a diagnostic test. The biomarkers include consistent immune and hormonal differences between long COVID and non-infected patients, including an "exaggerated" humoral immune response to the coronavirus and lower concentrations of cortisol among the former. The researchers developed a diagnostic algorithm to factor in these findings using machine learning, which was 96% accurate in distinguishing between long COVID patients and controls. Mount Sinai Health System's David Putrino called this development "a decisive step forward in the development of valid and reliable blood testing protocols for long COVID."
|
Negativity Bias Boosted Voter Fraud Conspiracy Theories Online
Futurity.org
Gregory Filiano
September 25, 2023
An international team of researchers led by Stony Brook University's Mason Youngblood found a negativity bias caused a surge of voter-fraud conspiracy theories on Twitter (now X) during the 2020 presidential election. The researchers modeled about 350,000 actual Twitter users' behaviors and correlated the sharing patterns of roughly 4 million voter fraud-related tweets with people more likely to retweet social posts containing stronger negative emotion. They learned these results align with previous research suggesting social media gives emotionally negative content an edge across diverse domains. Youngblood said the study’s result “has important implications for current debates on how to counter the spread of conspiracy theories and misinformation on social media.”
|
French Authorities Get Apple Software Update After iPhone 12 Dispute
Reuters
Elizabeth Pineau
September 26, 2023; et al.
A source at France's digital ministry reported that the government has been given an iPhone 12 software update from Apple after sales of the phones were stopped in that nation earlier this month due to tests detecting higher-than-allowed radiation exposure levels. Apple said it would provide a software update that accommodated France's testing methods after previously contending that the iPhone 12 was certified as compliant with global standards by multiple international bodies. Belgium's industry regulator indicated that it also had requested the software update, but because it has been provided solely to France, it is pushing for action at the European level.
|
Method Helps AI Navigate 3D Space Using 2D Images
NC State University News
Matt Shipman
September 25, 2023
Scientists at North Carolina State University (NC State), the University of Central Florida, China-based open Internet platform Ant Group, and smartphone technology developer OPPO Seattle Research Center have trained artificial intelligence (AI) to extract three-dimensional (3D) information from two-dimensional (2D) images through a technique they call MonoXiver. Current extraction methods have the AI scan 2D images and enclose objects within eight-point "bounding boxes" to determine their size and position relative to other objects in the image. MonoXiver uses each bounding box as an anchor, then performs an analysis to generate secondary bounding boxes surrounding that anchor. The AI compares each secondary box's geometry and appearance to ascertain which box has best captured any "missing" pieces of the object, supporting a highly efficient top-down sampling process, according to NC State's Tianfu Wu. He said MonoXiver "significantly improved the performance" of three 2D-to-3D data extraction techniques.
|
One-Stop Implementation from Signal Detection to Processing
Pohang University of Science and Technology (South Korea)
September 20, 2023
A team of researchers from South Korea's Pohang University of Science and Technology (Postech) and Chungnam National University Sejong Hospital and the U.K.'s University of Cambridge has developed a neural sensor for detecting, amplifying, and processing bio-signals. The researchers inkjet-printed the sensors on an ultra-thin substrate and attached them to the cerebral cortex of rats, where they demonstrated their ability to record high-resolution brain-originating signals. Postech's Sungjune Jung said, "This technology empowers the seamless creation of patterns in desired areas, paving the way for future manufacturing of customized bio-signal measurement devices."
|
Four-Hundred-Pound NYPD Robot Gets Tryout in Times Square Subway Station
The New York Times
Jeffery C. Mays
September 23, 2023
The New York Police Department (NYPD) has deployed a nearly 400-pound "fully autonomous outdoor security robot" manufactured by security technology company Knightscope in Times Square's subway station as a test run. The K5 robot uses four cameras to record video but not sound, and will not use facial recognition; it will spend 14 days mapping the station, accompanied by a human officer from midnight to 6 a.m. The robot is expected to patrol the station's mezzanine after the test as a "mobile camera" that straphangers could use to summon help. Knightscope said the weatherproof K5 travels at a maximum speed of three miles per hour, can provide high-definition recordings in 360 degrees, and issues audio messages. Its deployment dovetails with New York City Mayor Eric Adams' push to enhance law enforcement with technology, though privacy advocates are doubtful.
*May Require Paid Registration
|
Getting Audio from Still Images, Silent Videos
Northeastern Global News
Cody Mello-Klein
September 25, 2023
A machine learning tool developed at Northeastern University can obtain audio from still images and muted videos. Using the Side Eye tool, which leverages image stabilization technology standard in most smartphone cameras, it is possible to determine the gender of someone speaking off camera and the exact words they said. Northeastern's Kevin Fu explained that the small springs holding a camera lens suspended in liquid experience microscopic vibrations and the light is bent almost imperceptibly when someone speaks near a camera lens. Taking advantage of the rolling shutter method of photography used by most smartphone cameras, the researchers can extract sonic frequencies from those vibrations. Side Eye produces muffled audio, but the use of machine learning and training on certain words and audio enables it to extract a substantial amount of information, said Fu.
|
Computational Design of Molecules with Desired Properties
University of Luxembourg
September 19, 2023
Researchers at the University of Luxembourg, Cornell University, and the U.S. Department of Energy's Argonne National Laboratory have demonstrated a "freedom of design" principle's ramifications for enabling the design of molecules with desired traits. The researchers discovered weakly correlated quantum-mechanical properties of small molecules, which "highlights an intrinsic flexibility ... in CCS [the chemical compound space], wherein there seems to be very few limitations preventing a molecule from simultaneously exhibiting any pair of properties or for many molecules sharing an array of properties," said Cornell's Robert DiStasio Jr. The researchers looked for molecules with both large molecular polarizability and electronic gap using Pareto multi-property optimization, uncovering a CCS pathway comprised of unanticipated molecules with common structural and/or compositional changes.
|
Algorithm Improves Assessment of Genetic Risks for Non-White Populations
Johns Hopkins Bloomberg School of Public Health
September 25, 2023
A genetic risk-scoring algorithm developed by a team led by researchers at Johns Hopkins Bloomberg School of Public Health and the National Cancer Institute aims to reduce health-care disparities. The algorithm, CT-SLEB, was applied to genetic studies covering more than 5 million people across diverse populations to produce genetic scores for 13 traits in five ancestry categories: European, African, Latino, East Asian, and South Asian. CT-SLEB was trained on data from 23andMe, the Global Lipids Genetics Consortium, the National Institutes of Health's All of Us research program, and UK Biobank. For non-European-ancestry populations, the model outperformed standard polygenic risk-scoring models based mainly on European-ancestry datasets. The researchers also found CT-SLEB produced more accurate genetic risk scores for African-ancestry populations and is computationally faster than the closest competing methods.
|
|
