Welcome to the March 25, 2020 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."
|
|
Folding@Home Network More Powerful Than World's Top 7 Supercomputers Combined
Tom's Hardware Paul Alcorn March 21, 2020
The Folding@Home distributed computing network is currently churning out 470 petaflops of raw computing power, which is more powerful than the collective computing muscle of world's top seven supercomputers, in a push to defeat the coronavirus pandemic. That compares to the 149 petaflops of sustained output generated by the world's fastest supercomputer, the Oak Ridge National Laboratory (ORNL)'s Summit system. ORNL announced two weeks ago that Summit had been enlisted in the fight against COVID-19. Folding@Home said the number of contributors in its fight against the pandemic has risen 1,200%.
|
Towards an Unhackable Quantum Internet
Harvard University John A. Paulson School of Engineering and Applied Sciences Leah Burrows March 23, 2020
Harvard University and Massachusetts Institute of Technology researchers have invented a technique for correcting signal loss in quantum-signal transmission, using a prototype quantum node that captures, stores, and entangles bits of quantum information. This is a key development toward a practical and unhackable quantum Internet, and a step toward realizing long-distance quantum networks. The researchers used silicon-vacancy color centers in diamonds, in effect integrating an individual color-center into a nanofabricated diamond cavity, which corrals information-bearing photons and forces their interaction with the single color-center. The device stores the quantum data for milliseconds—sufficient for information to be conveyed over thousands of kilometers—with electrodes embedded around the cavity delivering control signals to process and preserve the data in the memory. Harvard's Mikhail Lukin said, "This is the first system-level demonstration ... that shows clear quantum advantage to communicating information using quantum repeater nodes."
|
There Is a Racial Divide in Speech-Recognition Systems, Researchers Say
The New York Times Cade Metz March 23, 2020
A study by Stanford University researchers found that speech-recognition systems from Amazon, Apple, Google, IBM, and Microsoft exhibited racial bias, misidentifying 35% of words spoken by people who were black, compared to about 19% of words spoken by white people. The systems also deemed approximately 2% of audio snippets from white people unreadable, versus 20% of snippets from blacks. The gap was just as large when comparing identical phrases spoken by both black and white people, indicating the problem is rooted in how the systems are trained to recognize sound. The researchers suggested the companies are training their systems on data that is insufficiently ethnically diverse.
*May Require Paid Registration
|
Simulated 'Frankenfish Brain-Swaps' Reveal Senses Control Body Movement
New Jersey Institute of Technology Jesse Jenkins March 23, 2020
Researchers at the New Jersey Institute of Technology (NJIT), Johns Hopkins University, and Turkey's Hacettepe University and Middle East Technical University used computer simulations of the brains and bodies of glass knifefish to model "fish brain transplants" in an effort to understand why organisms with different-sized limbs and muscles can perform identical fine-motor tasks equally well. The researchers swapped models of the fishes' information processing and motor systems, and found the resulting "Frankenfish" compensated for the brain-body mismatch by relying on sensory feedback to resume control of fine-motor movements for swimming. The researchers think these findings reinforce the argument for future robotic designs that utilize robust sensory feedback control systems, which may better adjust to unpredictable events in their environment.
|
Portable AI Device Turns Coughing Sounds Into Health Data for Flu Forecasting
University of Massachusetts Amherst March 19, 2020
University of Massachusetts, Amherst (UMass Amherst) researchers Tauhidur Rahman and Forsad Al Hossain have created a portable, artificial intelligence (AI)-driven device that can analyze coughing and crowd size in real time to track trends in flu-like illnesses. The FluSense platform processes data from microphones and thermal imaging using a Raspberry Pi and neural computing engine. FluSense devices in UMass's University Health Services clinic collected and analyzed more than 350,000 thermal images and 21 million non-speech audio samples from December 2018 to July 2019. The devices accurately predicted daily illness rates, while multiple and complementary sets of FluSense signals "strongly correlated" with laboratory testing for flu and flu-like illnesses.
|
Rapid Automatic Identification of Live Brain Cells
University of Tokyo (Japan) March 19, 2020
Researchers at the University of Tokyo in Japan have developed a computer algorithm to identify each neuron in fluorescent microscope images of live C. elegans roundworms. The researchers identified the neurons and quantified the distances and relative positions between neuron pairs in the images, then used these measurements and the C. elegans connectome brain atlas to develop the algorithm. The University of Tokyo's Yu Toyoshima said the algorithm is only 60% accurate, which is insufficient for fully automatic neuron identification, but "It speeds up our work enough to make other projects possible to understand neural networks based on whole-brain imaging data."
|
System Trains Driverless Cars in Simulation Before They Hit the Road
MIT News Rob Matheson March 23, 2020
Massachusetts Institute of Technology researchers have created a photorealistic simulation system to train driverless cars prior to their deployment on actual streets. The Virtual Image Synthesis and Transformation for Autonomy (VISTA) system employs a small dataset assembled from the experiences of actual drivers, to synthesize a virtually infinite number of pathways the vehicle could take in the real world. The controller is rewarded for the distance it travels without crashing, which requires it to learn how to safely reach a destination by itself, and navigate any situation it encounters. Tests showed that a controller trained via such reinforcement learning in simulation could be safely deployed into a driverless car and navigate through previously unseen streets, the researchers said.
|
Cybercriminals Take Advantage of Coronavirus
The Wall Street Journal Jenny Strasburg; Drew Hinshaw; Catherine Stupp March 24, 2020
Hackers are targeting critical healthcare systems already strained by the coronavirus pandemic, compromising computer networks and disrupting patient care. Flavio Aggio of the World Health Organization (WHO) said criminals have stepped up exploits, including ransomware attacks, against the agency. WHO and other global health organizations have been combating scams spreading misinformation about the coronavirus, or exploiting anxieties to breach corporate networks and work-from-home workforces. Experts said criminals are taking advantage of the pandemic to launch more targeted attacks against healthcare facilities, whose resources are focused on the coronavirus. A ransomware attack disabled the Champaign-Urbana Public Health District's website in Illinois earlier this month, and hackers also crippled a Czech Republic hospital's computer network.
*May Require Paid Registration
|
Researchers Use 3D Printing, Sensors to Create Models for Hydropower Testing
Oak Ridge National Laboratory March 20, 2020
Researchers at the Oak Ridge National Laboratory (ORNL) are using three-dimensional (3D) printers and sensors to create fake fish to help test the environmental impact of hydroelectric dams on fish populations. Fish passing through such dams encounter pressure changes, turbulence, and spinning blades on turbines that can cause injury. As part of its environmental assessment work for hydropower projects, ORNL developed a system to analyze the impact of turbine designs on high-value species. The researchers used 3D printers to create molds for models of various fish species, which they created from ballistic gel, and embedded sensors in the models to measure acceleration and G-forces in a turbine simulator. Said ORNL’s Ryan Saylor, "Mimicking biology is difficult. We’ve worked to get as close as we can to a good representation without the use of specialized equipment that could increase costs.”
|
AI-Powered Shoes Unlock Secrets of Your Sole
Stevens Institute of Technology March 17, 2020
Stevens Institute of Technology researchers have developed a smart insole powered by artificial intelligence (AI) that analyzes a person's gait in real time when worn in their shoe. Stevens Wearable Robotic Systems Laboratory director Damiano Zanotto and colleagues designed the SportSole to monitor its movement and spatial orientation with accelerometers and gyroscopes, while force sensors detect plantar pressure. The SportSole makes 500 readings each second, which Zanotto condenses to a few key features that are fed to an AI algorithm that extracts gait parameters accurate to within a few percentage points. The SportSole operates whether the wearer is walking or running, and does not require customization for individual users. Zanotto said, "We're achieving the same or better results [as high-end sensors] at a far lower cost, and that's a big deal when it comes to scaling this technology."
|
Digital Simulations Herald New Era of Intelligent Measurement
Discover March 25, 2020
Researchers at the University of Amsterdam in the Netherlands used a computer model to measure wind speed from the movements of a simulated flag. The researchers built a database of 14,000 flag simulations in which they varied not just wind speed, but also the virtual camera angle and distance from the flagpole, the lighting angle, and other parameters. They then taught a machine learning algorithm to recognize flags flying in the same wind speed, even when various other parameters differ. The algorithm analyzed a database of 4,000 real flag videos, and compared them to the simulations to recommend a change in the modeled parameters to improve realism. The results demonstrated how physical parameters captured in actual videos may be measured using simulations.
|
Windows Code-Execution Zeroday Is Under Active Exploit, Microsoft Warns
Ars Technica Dan Goodin March 23, 2020
Microsoft has issued a warning that a Windows zero-day vulnerability is being exploited in "limited targeted attacks" to execute malicious code on fully updated systems. The font-parsing remote code-execution vulnerability exists in the Adobe Type Manager Library, which numerous apps use to manage and render fonts available from Adobe Systems. The two code-execution flaws can be exploited by convincing a target to open or view a booby-trapped document in the Windows preview pane. Said Microsoft in an advisory, "For systems running supported versions of Windows 10, a successful attack could only result in code execution within an AppContainer sandbox context with limited privileges and capabilities." Until a patch is made available, Microsoft recommends disabling the Preview Pane and Details Pane in Windows Explorer, disabling the WebClient service, or renaming ATMFD.DLL or disabling the file from the registry.
|
Facebook Sent Home Thousands of Human Moderators Due to the Coronavirus. Now the Algorithms Are in Charge
The Washington Post Elizabeth Dwoskin; Nitasha Tiku March 24, 2020
Facebook recently placed its army of human moderators on paid leave, forcing the company to police disinformation without them. During the ongoing coronavirus pandemic, Facebook is relying more heavily on artificial intelligence software to make the subjective decisions about which content violates its terms of service and should be removed from the platform. Users should expect more mistakes while the company expedites the process, and there could be a rise in "false positives," including removal of content that should not be taken down, according to Facebook CEO Mark Zuckerberg. YouTube and Twitter also announced temporary plans to rely more heavily on automated systems. The decision to send moderators home and rely more on technology to police the sites concerned some researchers. Said Mary Gray at Microsoft Research, “They haven't made enough leaps and bounds in artificial intelligence to take away the best tool we have: human intelligence to do the discernment."
|
|