Welcome to the May 16, 2022, 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."
|
|
|
The Man Who Controls Computers with His Mind
The New York Times Magazine
Ferris Jabr
May 15, 2022
Paralyzed since 2006, Dennis DeGray has regained a semblance of control over his body via a brain-computer interface (BCI) developed by Stanford University researchers. Implanted in him in 2016, the BCI enables DeGray to move a cursor on a computer screen by thought, using machine learning algorithms that associate different neural activity patterns with different intended hand movements. DeGray has learned to control various technologies with his mind, including videogames, robotic limbs, and a simulated aerial drone. BCI advancements to date have relied on a combination of invasive and noninvasive technologies. Thomas Oxley at BCI developer Synchron believes future models will help physically disabled people re-engage with physical and digital environments.
|
Smart Tech Helping to Save China's Giant Pandas
CNN
Rebecca Cairns
May 16, 2022
China's Sichuan Forest and Grassland Administration and Chinese technology company Huawei jointly developed the Digital Panda System to protect the habitat of giant pandas by detecting wildfires in hard-to-reach regions where they live and notifying rangers and fire departments so they can intercede quickly. The solar-powered system collates data from 596 cameras, 45 infrared cameras, drones, and satellites, and stores it in the cloud. The system also incorporates a facial recognition system for pandas, based on an algorithm tested and refined using a database of over 6,400 images taken from 218 pandas in captivity.
|
AI Predicts Infant Age, Gender Based on Temperament
WSU Insider
Will Ferguson
May 10, 2022
A multi-institutional team of researchers used artificial intelligence to analyze temperament data on 4,438 infants in order to classify them by gender and age. The researchers utilized infant behavior data collected between 2006 and 2019, which rated infants along 14 temperament dimensions. The University of Idaho's Erich Seamon used machine learning algorithms to classify infants as either male or female at zero to 24 weeks, 24 to 48 weeks, and older than 48 weeks based on those ratings. The researchers determined the algorithms better classified gender in babies more than 48 weeks old, while fear was the most critical element in distinguishing boys and girls in the youngest and mid-range age groups.
|
More Efficiency for Optical Quantum Gates
Max Planck Institute of Quantum Optics (Germany)
May 12, 2022
Researchers at Germany's Max Planck Institute of Quantum Optics (MPQ) have created quantum gates that harness photons with greater efficiency. The MPQ's Stephan Dürr said, "We have succeeded for the first time in realizing an optical two-qubit [quantum bit] gate with an average efficiency of more than 40%." The MPQ's Thomas Stolz credited this achievement to non-linear components deployed in a novel experimental platform comprised of a low-density atomic gas cooled to 0.5 microkelvin, with atoms positioned between the mirrors of an optical resonator. Said Dürr, "Quantum gates for qubits implemented in trapped ions or superconducting materials are currently the most technically advanced,” adding, “realizing such an element with photons is much more challenging."
|
Python Is Slow, but About to Get Faster
ZDNet
Liam Tung
May 13, 2022
Core Python (CPython) developer Mark Shannon told the PyCon 2022 conference about the latest improvements in the Python programming language's speed developed by the Faster CPython Project. The project aims to double the speed of the current stable 3.10 series, and to improve Python performance levels to be closer to those of C. Results released ahead of the conference indicated the 3.11 beta preview was 1.25 times faster than version 3.10. Shannon thinks the enhancements can extend Python's utility to more virtual machines. He wrote in Python Enhancement Proposal 659 last year that the upgrade consists of a "specializing, adaptive interpreter that specializes code aggressively, but over a very small region, and is able to adjust to mis-specialization rapidly and at low cost."
|
Drones, ML Help Reveal the Scale of Plastic Pollution
University of Tsukuba (Japan)
May 12, 2022
Researchers at Japan's University of Tsukuba have combined flying drones and machine learning (ML) to map plastic pollution in river habitats. The researchers employed drones to capture high-resolution optical and thermal images along Indonesia's Brantas River, which were processed by ML algorithms. "Using these methods, we found that a combination of optical and thermal images produced the most accurate estimates of the amounts of plastic litter," said Tsukuba's Fatwa Ramdani. "On their own, neither image type produced particularly accurate results."
|
Technique Protects Privacy When Making Online Recommendations
MIT News
Adam Zewe
May 12, 2022
Massachusetts Institute of Technology (MIT) scientists have developed a method that shields personal data while ensuring the accuracy of online recommendations. The protocol relies on two separate servers accessing the same database, enhancing efficiency while facilitating private information retrieval; MIT's Sacha Servan-Schreiber said this allows clients to query a database without exposing the object of the search. The researchers also applied a tuning technique to eliminate many possible nearest neighbors for the client, along with an approach called oblivious masking to conceal any additional data points, apart from the actual nearest neighbor. The technique can also minimize data leakage, even if a bad actor attempts to fool a database into disclosing secret information.
|
Some Top 100,000 Websites Collect Everything You Type—Before You Hit Submit
Ars Technica
Lily Hay Newman
May 14, 2022
Researchers at Belgium's Katholieke Universiteit Leuven, Radboud University in the Netherlands, and Switzerland's University of Lausanne analyzed the top 100,000 websites and found a significant number record some or all of visitors' typewritten data. The researchers estimated 1,844 sites gathered a European Union user's email address without consent, while 2,950 logged a U.S. user's email. Many sites incorporate third-party marketing and analytics services that perform data-logging. After crawling sites for password leaks last May, the researchers found 52 sites in which third parties, including Russian technology company Yandex, were incidentally collecting password data prior to submission.
|
ML Improves Rainfall Mapping, Water Plans
University of Hawaii News
Maria Dumanlang
May 10, 2022
The U.S. National Science Foundation's Hawaii EPSCoR Ike Wai project funded University of Hawaii (UH) and East-West Center researchers' development of more accurate machine learning-based monthly rainfall maps. "This approach demonstrates how, with a moderate amount of data, a low-level machine learning algorithm can be used to train, evaluate, and classify an unrealistic map output," said UH's Matthew Lucas. Using UH's Mana high-performance computing cluster, the work yielded high-resolution, gridded monthly rainfall time series data for Hawaii spanning three decades. Said the East-West Center's Ryan Longman, "Having a wide range of gridded products will allow researchers the opportunity to develop important decision support for the state such as fire, flood, and drought risk and early warning systems."
|
Online Sentiment About Vaccines Previews Later Vaccination Rates
New York University
May 14, 2022
New York University researchers analyzing Twitter posts found that the sentiments they expressed about COVID-19 vaccines anticipated later vaccination rates. The researchers implemented a real-time big data analytics framework based on Natural Language Processing, Sentiment Analysis, and Amazon Web Services to review the posts. The system thematically classified vaccine-related tweets and catalogued them as positive, negative, or neutral. The researchers also used the IEEE Dataport dataset, which tagged tweets' coronavirus-related sentiment scores by U.S. geographic location. Positive sentiment toward vaccinations was followed by higher vaccination rates in the same region a week later; lower vaccination rates followed a week after negative sentiments were expressed in the same area.
|
Robot Bartender Won't Flirt with Your Date, but You Still Have to Tip
Hastings Tribune
Ron Hurtibise
May 13, 2022
Italian robotic bartending systems developer Makr Shakr said its Bionic Bar system has been installed aboard Royal Caribbean's Wonder of the Seas cruise ship. Bionic Bar's creators modeled the robots' movements after those of dancers, to make them seem less mechanical. Customers order drinks on a tablet, and the system’s two robotic arms draw precise measurements from roughly 140 liquor bottles and 16 mixers extended upside down above the bar. Makr Shakr said the system can make up to 250 drinks per hour, and patrons of the system will find it charges the same 18% gratuity charged at every other bar on Royal’s cruise ships.
|
|
