Welcome to the December 26, 2018 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."
|
|
|
A Robotic Hand Plays the Piano With a More Human Touch
The New York Times
Steph Yin
December 19, 2018
Researchers at the University of Cambridge in the U.K. have developed a robotic hand that can play the piano. The hand operates passively, meaning its fingers are not individually connected to any motors; instead, a simple mechanical arm controls only the wrist, and the rest of the hand follows using a design that mimics human anatomy. For now, the hand is only capable of performing simple songs, but it can execute different styles and dynamics. The researchers designed the hand with bones and ligaments placed as they are in nature. The team evaluated how hard or soft different components needed to be, and fabricated the hand with a three-dimensional printer that blended hard plastic and soft rubber in different ratios to create ligaments and joints with varying degrees of stiffness. Said Cambridge’s Josie Hughes, “We’re going beyond simply treating notes as a list of instructions. Piano playing is an art. We’re trying to introduce that complexity, depth, and artistry into our robot’s behavior.”
|
Room for Improvement? New Hotelier Tests an Algorithmic Pricing System
The Wall Street Journal
Craig Karmin
December 19, 2018
The Life House hotelier is experimenting with an algorithmic pricing system to set different room rates and customize promotions for guests based on personal information. Once customers book for the first time at a standard rate, they fill out a questionnaire to specify how often they travel, how frequently they visit the hotel bar, their number of Instagram followers, and other details. Customers who mention Life House on social media can enjoy discounts and other perks. In amassing information on guests, the algorithm is designed to refine a customer profile's value and other criteria to the hotelier, then adjust each guest's room rates and perks accordingly. PricewaterhouseCoopers' Scott Berman said, "Most of the big hotel operating companies are focused on their best guests," while Life House's algorithmic rate-setting "is next-generation."
*May Require Paid Registration
|
How AI Spotted Every Solar Panel in the U.S.
PBS NewsHour
Vicky Stein
December 19, 2018
Stanford University engineers have developed a method for locating every solar panel in the contiguous U.S. from a massive satellite image database via a deep learning computer model. The researchers used a pre-trained model called Inception as the basis for the DeepSolar neural network's clustering and classifying of pixels in images. DeepSolar scanned more than 1 billion image "tiles," comprising areas bigger than a neighborhood but smaller than a zip code; each tile had 102,400 pixels, and DeepSolar classified each pixel in each tile, determining whether it was likely part of a solar panel or not. The network completed this task in less than a month, ascertaining that regions with more sun exposure had greater solar panel adoption than areas with less average sunlight. DeepSolar also learned adoption was higher in locations of increasing average household income.
|
New Optical Memory Cell Achieves Record Data-Storage Density
Optical Society of America
December 20, 2018
Researchers at the universities of Oxford and Exeter in the U.K., and the University of Munster in Germany, have demonstrated a new method for storing more optical data in a smaller space on-chip than previously possible. The new technique enhances the phase-change optical memory cell, which uses light to write and read data, potentially yielding a faster, more power-efficient form of computer memory. The researchers used laser light with a single, double-stepped pulse to precisely adjust the melting and crystallization of the phase-change material. The device exhibited optical memory with 34 states, or levels, exceeding 5 bits; only 10 levels could be achieved previously. Oxford's Harish Bhaskaran said, "By bringing the speed of light-based data transmission to the circuit boards that run computers, our all-optical memory could enable a hybrid computer chip that interacts with data both optically and electrically."
|
An Amoeba Just Found an Entirely New Way to Solve a Classic Computing Problem
ScienceAlert
Fiona MacDonald
December 21, 2018
Researchers at Keio University in Japan gave the Traveling Salesman Problem (TPS) to a "true slime mold" amoeba, and found as the cities increased from four to eight, the single-celled organism only needed a linear amount of more time to determine a single reasonable route. TPS is an optimization problem requiring a computer to look at a list of cities and determine the shortest route in which each city is visited exactly once. In this experiment, the "cities" were 64 narrow channels—eight "cities" each containing eight channels—in a round plate placed on top of agar. The researchers ensured the amoeba entered the "cities" in an optimal way, using light to illuminate certain channels that were too far apart or which it had already visited, and to stop it from entering several channels simultaneously. The team said their results "may lead to the development of novel analog computers enabling approximate solutions of complex optimization problems in linear time."
|
Microsoft Collaborates With CodeJIKA to Offer Coding Skills
ITWeb
Sibahle Malinga
December 21, 2018
Microsoft has partnered with CodeJIKA, a nonprofit information and communications technology skills development organization, to establish coding clubs and teacher induction workshops in 60 schools across five provinces in South Africa. CodeJIKA is part of Code for Change, a global nonprofit that aims to provide young people with coding and computer-based skills to develop their computational and critical thinking, while showing them how to create and use new technologies. CodeJIKA hopes to have at least 20% of South African secondary schools teaching coding and computer science skills by 2020. Said Microsoft South Africa philanthropies lead Charlene Verzmoter, "We believe the real value in learning computer science is not in the act of coding itself, but in building students' ability to be creative, resilient and collaborative. These are skills they will need to succeed in the future."
|
Programming Proteins to Pair Precisely
UW News
Walter Neary
December 19, 2018
University of Washington (UW) scientists have designed proteins to pair up in much the same way DNA molecules bind into a double helix, potentially enabling protein nanomachines that help diagnose and treat disease and permit more precise cellular engineering. Said UW's Zibo Chen, "What we're doing is computationally designing these hydrogen-bond networks so that each protein pair has a unique complementary sequence. There is only one way for them to come together and they do not cross-react with proteins from other pairs." The Rosetta protein design algorithm leverages the fact that the configuration of an amino acid chain is shaped by attraction and repulsion between the amino acids of the chain and the fluid in which the chain is immersed. Rosetta calculates the shape that best balances out these forces for the chain to achieve its lowest overall energy level.
|
Is Genocide Predictable? Researchers Say Absolutely
National Public Radio
Jason Beaubien
December 20, 2018
Researchers at Dartmouth University and the U.S. Holocaust Memorial Museum have compiled a database of every mass killing since World War II, and probed the conditions in nations where the atrocities occurred shortly before they began. Using this computer model, the researchers are analyzing which countries are currently at the highest risk of genocide. The model studies statistics suspected to signify inequality, poverty, and economic imbalance, like fluctuating per capita gross domestic product, infant mortality rates, and populace. Recent coup attempts, authoritarianism levels, civil rights, political killings, and ethnic polarization also are factored in. The Holocaust Museum's Jill Savitt said, "What we're doing is trying to alert policymakers that here's a situation that is ripe for horrors to happen and give them a heads-up that there are actions that can be taken to avert it."
|
Companies Use VR to Train Employees for Hostage Situations, Robberies
CBS News
December 20, 2018
Major corporations like Walmart, Chipotle, and Verizon are using virtual reality (VR) to prepare employees for potentially dangerous situations on the job. For example, Verizon has more than 1,600 stores in the U.S. whose front-line employees participate in a digital scenario in which two gunmen strike as the store opens, taking one employee hostage and trying to break into the safe. The training features three virtual reality experiences: an armed robbery at a store opening, an armed robbery at closing, and a smash-and-grab job during store hours. Throughout the VR simulations, employees answer questions about what to do in real time, then discuss as a group. Verizon security chief Michael Mason said the goal of the program isn't to protect the stores’ inventory, but to protect the employees.
|
Machine Vision Can Create Harry Potter-Style Photos for Muggles
Technology Review
December 21, 2018
University of Washington in Seattle researchers have developed software that can animate the central character in a photograph while leaving the rest of the image untouched. The researchers used a program called SMPL from a team at Microsoft and the Max Planck Institute for Intelligent Systems in Germany, which begins with a two-dimensional (2D) cutout of a human body and superimposes a three-dimensional skeleton onto the shape. The skeleton can be animated to create the sense of movement, solving the problem of pose estimation for a limited set of circumstances. The code needs to see a head-to-toe cutout of a body viewed from the front, and while it can handle some types of occlusions, it cannot handle more complex occlusions. The researchers solved this problem by mapping a body-shaped mesh into 2D space and aligning it with the cutout using a warping algorithm; this identified specific parts of the body and warped them to match the cutout.
|
Real-Time Readouts of Thinking in Rats
MIT News
David Orenstein
December 19, 2018
Massachusetts Institute of Technology (MIT) researchers have developed an open source system delivering fast, precise neural decoding and real-time readouts of where rats think they are. The neural decoding software decrypts hippocampal spatiotemporal patterns detected from tetrode recordings without requiring spike sorting, an error-prone computational process. Implementing this software on a graphical processing unit (GPU) chip demonstrated a 20- to 50-fold upgrade in decoding and analysis speed over conventional multicore central processing unit (CPU) chips. The system also retained its speed and accuracy when accommodating more than 1,000 input channels, extending the real-time decoding approach to new high-density brain recording devices. Said New York University's Zhe Chen, formerly of MIT, "We are proposing an elegant solution using GPU computing to not only decode information on the fly, but also to evaluate the significance of the information on the fly."
|
New Attack Intercepts Keystrokes via Graphics Libraries
ZDNet
Catalin Cimpanu
December 19, 2018
A team of researchers from the University of California, Riverside, Virginia Polytechnic Institute and State University, and the U.S. Army Research Laboratory has developed a side-channel attack for intercepting keystrokes by exploiting the central-processing unit's shared memory, where graphics libraries manage rendering the operating system user interface. The attack code needs no admin/root or other special privileges to operate, and can be concealed within legitimate apps and retrieve keystrokes with less chance of detection by antivirus products. In comparison with classic keylogging malware, the attack also works with onscreen keyboards and is not restricted to collecting key presses from hardware keyboards. The exploit also can be modified for any operating system (OS), including mobile OSes. The attack becomes more adept at guessing correct key presses as it amasses more key presses from graphics libraries.
|
Brazilian Museum Destroyed by Fire Lives on Through Google
NBC News
Alyssa Newcomb
December 19, 2018
Google is helping preserve Brazil's National Museum via a virtual exhibition, three months after the institution's destruction by fire. The virtual exhibition is is the result of a venture the museum began two years ago to digitize its collection of 20 million artifacts as a virtual tour using Google's street view technology. Google Arts & Culture's Chance Coughenour said, "Advances in technology—like high-resolution photography, photogrammetry, [three-dimensional] laser scanning, and virtual and augmented reality—have not only introduced new forms of art, but help us preserve the world's most precious heritage. Even though images cannot replace what has been lost, they offer us a way to remember."
|
Scientists to Give AI Human Hearing
R&D Magazine
December 19, 2018
Researchers at Peter the Great St. Petersburg Polytechnic University (SPbPU) in Russia have simulated the process of sensory sound coding by modeling the mammalian auditory periphery. The researchers developed methods for acoustic signal recognition based on peripheral coding. The team partially reproduced the processes performed by the nervous system while processing information and integrated the process into a decision-making module, which determines the type of incoming signal. The SPbPU researchers believe this method could help create a new generation of neurocomputer interfaces, as well as provide better human-machine interaction. Said SPbPU's Anton Yakovenko, "The algorithms for processing and analyzing big data implemented within the research framework are universal and can be implemented to solve the tasks that are not related to acoustic signal processing."
|
|
