A Google Brain and Imperial College London team have built a system -- Pre-training with Extracted Gap-sentences for Abstractive SUmmarization Sequence-to-sequence, or Pegasus -- that leverages Google's Transformers architecture combined with pretraining objectives tailored for abstractive text generation. From a report: They say it achieves state-of-the-art results in 12 summarization tasks spanning news, science, stories, instructions, emails, patents, and legislative bills, and that it shows "surprising" performance on low-resource summarization, surpassing previous top results on six data sets with only 1,000 examples. As the researchers point out, text summarization aims to generate accurate and concise summaries from input documents, in contrast to executive techniques. Rather than merely copy fragments from the input, abstractive summarization might produce novel words or cover principal information such that the output remains linguistically fluent. Transformers are a type of neural architecture introduced in a paper by researchers at Google Brain, Google's AI research division. As do all deep neural networks, they contain functions (neurons) arranged in interconnected layers that transmit signals from input data and slowly adjust the synaptic strength (weights) of each connection -- that's how all AI models extract features and learn to make predictions. But Transformers uniquely have attention. Every output element is connected to every input element, and the weightings between them are calculated dynamically.

Read more of this story at Slashdot.

Experts say Saudi prosecution of Jamal Khashoggi's killers violated human rights, call for international investigation.

Government says the troops were killed after heavily armed fighters ambushed their convoy.

Thousands of Eritreans leave every year, with football players also 'defecting' while playing abroad.

Games have long served as a training ground for AI algorithms, and not without good reason. From a report: Games -- particularly video games -- provide challenging environments against which to benchmark autonomous systems. In 2013, a team of researchers introduced the Arcade Learning Environment, a collection of over 55 Atari 2600 games designed to test a broad range of AI techniques. More recently, San Francisco research firm OpenAI detailed Procgen Benchmark, a set of 16 virtual worlds that measure how quickly models learn generalizable skills. The next frontier might be Mega Man, if an international team of researchers have their way. In a newly published paper [PDF] on the preprint server Arxiv.org, they propose EvoMan, a game-playing competition based on the eight boss fights in Capcom's cult classic Mega Man 2. As they describe it, competitors' goal is to train an AI agent to defeat every enemy and evaluate their performances by common metrics. Why Mega Man 2? The paper's coauthors assert that few other environments test an AI's ability to win against a single enemy, or how well an AI generalizes to win matches against waves of enemies or coevolves to create increasingly difficult enemies. To this end, in EvoMan, an AI-controlled Mega Man -- a robot equipped with a powerful arm cannon -- must beat eight so-called Robot Masters equipped with different weapons. Every time a Robot Master is defeated, the agent acquires its weapon, making it easier to defeat the bosses who remain. As proposed, EvoMan challenge entrants would train their agents on any four enemies and measure how well their learned strategy scales up to a whole set of enemies. The agents would be expected to learn how to identify and react to general patterns like avoiding being shot or shooting at the direction of the enemy, and to deplete an enemy's health from 100 energy points to 0 by the end of each match.

Read more of this story at Slashdot.

Kyle Wiggers, writing for VentureBeat: Optical character recognition (OCR), or the conversion of images of handwritten or printed text into machine-readable text, is a science that dates back to the early '70s. But algorithms have long struggled to make out characters that aren't parallel with horizontal planes, which is why researchers at Amazon developed what they call TextTubes. They're detectors for curved text in natural images that model said text as tubes around their medial (middle) axes. In a paper [PDF] describing their work, the coauthors claim that their approach achieves state-of-the-art results on a popular OCR benchmark. As the researchers explain, scene text is typically broken down into two successive tasks: text detection and text recognition. The first involves localizing characters, words, and lines using contextual clues, while the second aims to transcribe their content. Both are easier said than done -- text in the wild is affected not only by deformations, but by viewpoint changes and arbitrary fonts. The team's solution is a "tube" representation of the text reference frame that captures most of the variability, taking advantage of the fact that target text is usually a concatenation of characters of similar size. It's formulated as a mathematical function that enables the training of machine learning scene text detectors, in contrast to traditional approaches that use overlap- and noise-prone rectangles and quadrilaterals to capture text information.

Read more of this story at Slashdot.

Saudi Arabia to launch investigation after admitting to carrying out an attack in Saada on Tuesday.

In a rare bipartisan achievement, Congress has moved to combat the scourge of robocalls inundating Americans. Just don't expect the phone to stop ringing any time soon. From a report: Lawmakers, industry and consumer groups say the bill represents significant steps forward, but they also concede that the calls are likely to continue -- a reflection of how a lasting solution continues to elude the companies and regulators that control the telephone system. "This isn't going to eliminate every robocall," said Sen. John Thune (R., S.D.), one of the bill's prime sponsors, in an interview. "But we think it will go a long way toward getting at some of these not only annoying nuisance calls, but more importantly a lot of scam artists that prey on vulnerable populations." Even the new law's name -- the TRACED Act, for Telephone Robocall Abuse Criminal Enforcement and Deterrence -- makes clear the goal is to deter robocallers rather than eradicate them. Lawmakers have previously tried and failed to stop robocalls, most prominently with the opening of the National Do Not Call Registry about 15 years ago. (Criminals ignore the list of off-limits numbers.) Implementing this latest anti-robocall law is likely to take years, telecom industry executives and robocall experts say. New consumer-protection rules will take months to craft and longer to implement. Lawmakers also left aspects of the problem for future study, calling for eight new reports and two new working groups. Some of the billions of robocalls Americans receive are legitimate, such as calls from a pharmacy telling a customer a prescription is ready. The calls are illegal when used for scams, or when they violate consumer-protection rules such as those against calling someone without their permission using a recorded message.

Read more of this story at Slashdot.

U.S. consumers spent more online this holiday season than they ever have before, according to a new report from Mastercard. From a report: Digging into the report, Reuters found e-commerce sales contributed 14.6 percent of total retail and were up 18.8 percent from 2018. This is despite a later Thanksgiving, which cut six days from the festive shopping window. Physical retailers are under increasing pressure from Amazon, which continues to offer ever more attractive ways to shop from the sofa, including same-day delivery in some cities and free returns on an ever-widening catalog of goods. The online retailer is so aggressive in its efforts to speed up delivery, in fact, it's taking a worrying toll on those making the deliveries.

Read more of this story at Slashdot.

Recently established political party Lelo prepares for parliamentary elections as country is gripped by protests.

The Federal Aviation Administration put forward a rule Thursday that would empower the government to track most drones in the U.S. From a report: The rule will require drones to implement a remote ID system, which will make it possible for third parties to track them. The measure will help law enforcement identify unauthorized drones that may pose a security threat, paving the way for wider adoption of commercial drone technology. The rule said that the FAA expects all eligible drones in the U.S. to comply with the rule within three years. The approval is a milestone in commercial drone delivery, as companies including Amazon, Uber and Google parent Alphabet are racing to add unmanned aircraft to their fleets to save costs and deliver goods faster. In June, Amazon debuted its newest delivery drone as part of a push inside Amazon to speed up its delivery times for Prime members. In October, Alphabet's drone unit Wing officially launched the country's first commercial drone delivery flight. UPS's Flight Forward subsidiary said in October that it received federal approval to operate a fleet of drones, giving it broad privileges to expand unmanned package delivery. It was the first time the FAA had granted such broad approval to a company to operate a fleet of drones as an airline.

Read more of this story at Slashdot.

At least two tech review sites are discussing whether to rescind their positive recommendations of Ring's home surveillance cameras, a leading digital-rights organization announced this week. From a report: In the wake of reporting by Gizmodo and other outlets this year concerning Ring's troubled security and privacy practices, Fight for the Future has launched a campaign calling on tech review sites, such as Consumer Reports and PC Magazine, to suspend recommending Ring products. "Tech reviews and guides play an important role in people deciding which devices to buy," said Evan Greer, deputy director of Fight for the Future. [...] Last week, the tech review site Wirecutter announced it was suspending its recommendation of Ring products citing a report about a data leak by BuzzFeed's Caroline Haskins. This prompted Fight for the Future to contact other review sites and ask them to rescind their recommendations as well.

Read more of this story at Slashdot.

Turkey's Constitutional Court ruling opens the way for the ban on online encyclopaedia to lifted.

A brief tour of some of the best literature from the Middle East and North Africa in 2019.

We’ve been through this drill before, with the Kavanaugh nomination. The Republicans have this angst filled “debate“ and in the end they do the evil thing. It's a trick. They find they can get credit for being open-minded and fair without actually having to be open-minded and fair.