Welcome to the July 28, 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."

Erriyon Knighton and Noah Lyles compete in the Men's 200 Meter Final of the 2020 U.S. Olympic Track & Field Team trials held in June in Eugene, OR. How Olympic Tracking Systems Capture Athletic Performances
Scientific American
Eleanor Cummins
July 27, 2021

This year's Olympic Games in Tokyo use an advanced three-dimensional (3D) tracking system that captures athletes' performances in fine detail. Intel's 3DAT system sends live camera footage to the cloud, where artificial intelligence (AI) uses deep learning to analyze an athlete's movements and identify key performance traits like top speed and deceleration. 3DAT shares this information with viewers as slow-motion graphic representations of the action in less than 30 seconds. Intel's Jonathan Lee and colleagues trained the AI on recorded footage of elite track and field athletes, with all body parts annotated; the model could then link the video to a simplified rendering of an athlete's form. The AI can track this "skeleton" and calculate the position of each athlete's body in three dimensions as it moves through an event.

Full Article
Gaming Graphics Card Allows Faster, More Precise Control of Fusion Energy Experiments
University of Washington News
Sarah McQuate
July 22, 2021

University of Washington (UW) scientists formulated a technique for using a gaming graphics card to control plasma formation in an experimental fusion reactor. The team utilized a Tesla graphics processing unit (GPU) from NVIDIA, which is engineered for machine learning applications. The card enabled the team to refine how plasmas entered the reactor, offering a more precise view of plasma formation. The prototype reactor self-generates magnetic fields within the plasma, making it potentially smaller and more affordable than other reactors that employ external fields. UW's Chris Hansen said, "This new system lets us try newer, more advanced algorithms that could enable significantly better control, which can open a world of new applications for plasma and fusion technology."

Full Article

A wearable device can reduce collisions and falls in the visually impaired. Wearable Camera Reduces Collision Risk for Blind, Visually Impaired
Brian P. Dunleavy
July 22, 2021

A wearable computer vision device developed by Harvard Medical School scientists may help reduce collisions and other accidents for the blind and visually impaired. The device includes a data recording unit enclosed in a sling backpack with a camera on the strap, and two Bluetooth-connected wristbands. The researchers said a processing unit records images from the camera and analyzes collision risk based on the motion of incoming and surrounding objects within the field of view. The left-hand or right-hand wristband will vibrate an alert depending on which side a potential collision is coming from, while both wristbands vibrate when a potential head-on collision is detected. Data from the study showed the solution cut the risk for collisions and falls by nearly 40% compared with other mobility aids, when used in combination with a long cane or guide dog.

Full Article

Cassie the robot completing the 5-kilometer route in just over 53 minutes. Bipedal Robot Learns to Run, Completes 5K
Oregon State University News
Steve Lundeberg
July 25, 2021

An untethered bipedal robot completed a five-kilometer (3.10-mile) run in just over 53 minutes. The Cassie robot, engineered by Oregon State University (OSU) researchers and built by OSU spinout company Agility Robotics, is the first bipedal robot to use machine learning to maintain a running gait on outdoor terrain. The robot taught itself to run using a reinforcement learning algorithm, and it makes subtle adjustments to remain upright while in motion. OSU's Jonathan Hurst said Cassie's developers "combined expertise from biomechanics and existing robot control approaches with new machine learning tools.” Hurst added, “In the not-very-distant future, everyone will see and interact with robots in many places in their everyday lives, robots that work alongside us and improve our quality of life.”

Full Article
Lost in L.A.? Fire Department Can Find You with What3words Location Technology
Stephen Shankland
July 22, 2021

The Los Angeles Fire Department (LAFD) has entered into a partnership with digital location startup What3words, which assigns a unique three-word name to each of 57 billion 10-foot-square spots on Earth. The department had been testing the application since last year, using it to locate places that emergency crews needed to reach even if the sites lacked conventional addresses. LAFD receives What3words locations through 911 calls on Android phones or iPhones, or through text messages sent by dispatchers with links that retrieve the three-word addresses. People also can use the What3words app to pinpoint their own locations. Increasing numbers of signs identify locations with their What3words designations, particularly in wildlands.

Full Article
High-Speed Projectors Power Virtual Air Hockey With Shape-Changing Paddles
Gizmodo Australia
Andrew Liszewski
July 21, 2021

Scientists at Japan's Tohoku University have invented a virtual version of air hockey that uses shapeshifting virtual projections of paddles and pucks. The MetamorHockey system replaces the traditional air hockey table surface with a semi-transparent rear-projection screen, which enables an underside projection to show through while a video camera underneath tracks the movements of each player's paddle. The paddle has an infrared light-emitting diode to facilitate tracking of its position and orientation. Both projector and camera operate at 420 frames per second; this data is fed to a computer that calculates the puck's movements and trajectories, then passes the information to a projector that refreshes the virtual objects' positions.

Full Article

Teeth, a bar in San Francisco, uses QR code technology for its menus. QR Codes Are Here to Stay. So Is the Tracking They Allow.
The New York Times
Erin Woo
July 26, 2021

Quick response (QR) codes that facilitate touchless transactions have become a permanent part of life, adopted by many varieties of commercial establishments. QR codes can store digital data including when, where, and how often a code-scan occurs, enabling businesses to integrate more tracking, targeting, and analytics tools. The American Civil Liberties Union's Jay Stanley said, "Suddenly your offline activity of sitting down for a meal has become part of the online advertising empire." Author Scott Stratten said the U.S. adoption of QR codes surged as a result of Apple enabling iPhone cameras to recognize the codes in 2017, and the coronavirus pandemic. Stanford University's Lucy Bernholz said she considers QR codes "an important first step toward making your experience in physical space outside of your home feel just like being tracked by Google on your screen."

Full Article
*May Require Paid Registration
Flexible Computer Processor Is Most Powerful Plastic Chip Yet
New Scientist
Matthew Sparkes
July 21, 2021

The newest processor from U.K. chip designer Arm reportedly can be printed directly onto paper, cardboard, or cloth. Arm's James Myers said the 32-bit PlasticARM chip can run various applications, although it presently uses read-only memory, and so can only execute the code with which it was built. The processor features circuits and components printed onto a plastic substrate, with 56,340 elements taking up less than 60 square millimeters. PlasticARM has roughly 12 times as many components to conduct calculations as the previous best flexible chip, with the potential to give everyday items like clothing and food containers the ability to collect, process, and transmit information across the Internet.

Full Article
Companies Beef Up AI Models with Synthetic Data
The Wall Street Journal
Sara Castellanos
July 23, 2021

Companies are building synthetic datasets when real-world data is unavailable to train artificial intelligence (AI) models to identify anomalies. Dmitry Efimov at American Express (Amex) said researchers have spent several years researching synthetic data in order to enhance the credit-card company's AI-based fraud-detection models. Amex is experimenting with generative adversarial networks to produce synthetic data on rare fraud patterns, which then can be applied to augment an existing dataset of fraud behaviors to improve general AI-based fraud-detection models. Efimov said one AI model is used to generate new data, while a second model attempts to determine the data's authenticity. Efimov said early tests have demonstrated that the synthetic data improves the AI-based model's ability to identify specific types of fraud.

Full Article
*May Require Paid Registration
Extracting More Accurate Data From Images Degraded by Rain, Nighttime, Crowded Conditions
Yale-NUS College (Singapore)
July 19, 2021

Novel computer vision and human pose estimation methods can extract more accurate data from videos obscured by visibility issues and crowding, according to an international team of scientists led by researchers at the Yale-National University of Singapore College. The research team used two deep learning algorithms to enhance the quality of videos taken at night and in rainy conditions. One algorithm boosts brightness while simultaneously suppressing noise and light effects to produce clear nighttime images, while the other algorithm applies frame alignment and depth estimation to eliminate rain streaks and the rain veiling effect. The team also developed a technique for estimating three-dimensional human poses in videos of crowded environments more reliably by combining top-down and bottom-up approaches.

Full Article
Framework Applies ML to Atomistic Modeling
Northwestern University McCormick School of Engineering
Alex Gerage
July 21, 2021

A new framework uses machine learning to enhance the modeling of interatomic potentials—the rules governing atomic interaction—which could lead to more accurate predictions of atomic-level nanomaterial behavior. An international team led by researchers from Northwestern University’s McCormick School of Engineering designed the framework, which applies multi-objective genetic algorithm optimization and statistical analysis to minimize user intervention. Northwestern's Horacio Espinosa said the algorithms "provide the means to tailor the parameterization to applications of interest." The team found the accuracy of interatomic potential correlated with the complexity and number of the stated parameters. Said Northwestern's Xu Zhang, "We hope to make a step forward by making the simulation techniques more accurately reflect the property of materials."

Full Article

The SentinelOne printer security bug has the potential to affect over 400 different printer models. Ancient Printer Security Bug Affects Millions of Devices Worldwide
Mayank Sharma
July 21, 2021

Cybersecurity researchers at SentinelOne have identified a highly severe privilege escalation vulnerability in HP, Samsung, and Xerox printer drivers. The vulnerability appears to have been present since 2005. The researchers said millions of devices and users worldwide likely have been impacted by the buffer overflow vulnerability, which can be exploited whether or not a printer is connected to a targeted device. SentinelOne's Asaf Amir said, "Successfully exploiting a driver vulnerability might allow attackers to potentially install programs; view, change, encrypt, or delete data, or create new accounts with full user rights." Hackers would need local user access to the system to access the affected driver and take advantage of the vulnerability.

Full Article
Big Data-Derived Tool Facilitates Closer Monitoring of Recovery From Natural Disasters
Texas A&M Today
Vandana Suresh
July 22, 2021

A framework based on location-based data on visits to essential establishments is designed to monitor communities' resilience to natural disasters. Texas A&M University (TAMU) researchers partnered with location data provider SafeGraph to access community-level big data gleaned by companies that monitor visits to locations within a perimeter based on anonymized cellphone data. The researchers first found points of interest corresponding to locations of establishments like hospitals, gas stations, and stores in Harris County, TX, that saw visitor traffic change after Hurricane Harvey made landfall. They then sifted through the data and counted the number of visits to each point of interest before and during the hurricane. TAMU's Ali Mostafavi said, "We'd like to create an intelligent dashboard that would display the rate of recovery and impacts in different areas in near-real time and also predict the likelihood of future access disruption and recovery patterns after a heavy downpour."

Full Article
The VR Book - Human-Centered Design for Virtual Reality
ACM Queue Case Studies

Association for Computing Machinery

1601 Broadway, 10th Floor
New York, NY 10019-7434

ACM Media Sales

If you are interested in advertising in ACM TechNews or other ACM publications, please contact ACM Media Sales or (212) 626-0686, or visit ACM Media for more information.

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