AI News, KCT artificial intelligence

This Picasso painting had never been seen before. Until a neural network painted it.

Art historians have long noted the presence of a ghostly woman’s face faintly visible beneath the paint.

In 1998, conservators at the Art Institute of Chicago, where the painting hangs, photographed it using x-rays and infrared light to see what lies beneath the surface.

Infrared and x-ray images show only the faintest outlines, and while they can be used to infer the amount of paint the artist used, they do not show color or style.

Anthony Bourached and George Cann at University College London have used a machine vision technique called neural style transfer to retrieve the lost Picasso painting in color for the first time.

The researchers have used the same technique to retrieve lost paintings by other artists and say it has the potential to transform the way art historians work.

The first layer might recognize broad features like edges, the next layer sees how these edges form simple shapes like circles, the next layer recognizes patterns of shapes, such as two circles close together, and yet another layer might label these pairs of circles as eyes.

They have taken a manually edited version of the x-ray images of the ghostly woman beneath The Old Guitarist and passed it through a neural style transfer network.

“We present a novel method of reconstructing lost artwork, by applying neural style transfer to x-radiographs of artwork with secondary interior artwork beneath a primary exterior, so as to reconstruct lost artwork,” they say.

“Our method of combining original but hidden artwork, subjective human input, and neural style transfer helps to broaden an insight into an artist’s creative process,” they say.

AI 50: America’s Most Promising Artificial Intelligence Companies

Artificial intelligence is infiltrating every industry, allowing vehicles to navigate without drivers, assisting doctors with medical diagnoses, and mimicking the way humans speak.

As Jeremy Achin, CEO of newly minted unicorn DataRobot, puts it: “Everyone knows you have to have machine learning in your story or you’re not sexy.” The inherently broad term gets bandied about so often that it can start to feel meaningless and gets trotted out by companies to gussy up even simple data analysis.

To help cut through the noise, Forbes and data partner Meritech Capital put together a list of private, U.S.-based companies that are wielding some subset of artificial intelligence in a meaningful way and demonstrating real business potential from doing so.

To be included on the list, companies needed to show that techniques like machine learning (where systems learn from data to improve on tasks), natural language processing (which enables programs to “understand” written or spoken language), or computer vision (which relates to how machines “see”) are a core part of their business model and future success.

Only eight startups were founded or cofounded by women, reflecting trends in venture funding, where software startups run by men have received the lion’s share of investment dollars.

Its software cross-references CT images of a patient’s brain with its database of scans and can alert specialists in minutes to early signs of large vessel occlusion strokes that they may have otherwise missed or taken too long to spot.

CEO Wout Brusselaers says the company’s software can pull data from electronic medical records to create patient graphs that allow researchers to filter for specific conditions and traits, leading to matches in “minutes, instead of months.” The system’s language understanding engine has been trained so that it can infer some conditions even if they’re not explicitly mentioned in notes, and Deep 6 says it has more than 20 health system or pharmaceutical customers.

CEO and cofounder Dhananjay Sampath launched Armorblox into the saturated cybersecurity market two years ago with the aim of protecting customers from socially engineered attacks, like phishing emails, that take advantage of human missteps.

“However, the real-world use cases tend to move in the opposite direction—demanding solutions with less compute, memory and power.” They set about trying to create a system where complex algorithms could run on simple hardware and spun out of the Allen Institute, which was cofounded by the late Paul Allen of Microsoft, in 2016.

The company admits that its AI agent, Chloe, is still in its infant stage right now—it can complete simple tasks like reading the instructions on a pill bottle — but it has big ambitions for more robust computer-vision-based navigation.

While chat bots burst onto the scene with a lot of promise (remember how they were going to take over Facebook Messenger?), they never quite reached mainstream adoption, due in part to disenchantment with their limited scope and conversational rigidity.

it just happens to be stuck in a research paper somewhere.” He and longtime Microsoft employee Diego Oppenheimer banded together to devise an easier way for data scientists to discover and work with machine learning models.

The app can build a 3D map of a space in roughly 30 seconds and uses computer vision to recognize real-world objects, so that objects created in AR can interact with them like they would in the real world.

The company recently raised a fresh round of funding led by automotive company Aptiv with the hope that its technology could one day be integrated into smart cars (imagine a vehicle that could issue a warning to a drowsy-looking driver).In the meantime, it’s also being used to test consumer feedback on ads and TV programming.

Blue Hexagon, led by longtime Qualcomm executive Nayeem Islam, spent more than a year and a half building a deep learning system to analyze network traffic that it says can detect and block threats in under a second.

It evaluates data from hundreds of online and offline data sources including credit bureaus, carrier phone records, IP addresses, social networks and more, to monitor for any suspicious behavior.

Chief technology officer Sarjoun Skaff says it has taken iteration after iteration since 2013 to figure out how to let its robots maneuver safely around shoppers and interpret billions of images in a way that was accurate, timely and reliable.

Pymetrics gleans key emotional and cognitive traits for different roles so when job seekers apply to work at one of those companies and complete the challenges themselves, they’re paired with jobs that are the best fit.

The company’s consumer app draws on a data set of more than 2 billion anonymized medical records, finding subtle patterns in the data to give users personalized health advice.

In July, K announced a partnership with insurance provider Anthem to let members see how doctors diagnose and treat similar people with similar symptoms for free (though they’ll be charged to chat with an actual doctor).

Computers were too slow and data too expensive to make AI practical at the time, but roughly three decades later, Pratt teamed up with investment firm TPG to help it identify a data analytics company to buy or invest in.

The company uses vast quantities of nontraditional data—like signals from cellphones, internet-of-things devices and street cameras—to issue hyperlocal “street-by-street, minute-by-minute” weather forecasts.

Its image processing technology, called Cortex, works with its own 3D camera, as well as a selection ofcheaper360-degree cameras, to let users create virtual versions of their space.The company’s leaning into the real estate market, showcasing how agents can use it to give 3D tours.

Beck, who spent more than five years as a pathologist at Harvard, wants to make it easier for other pathologists to diagnose diseases like cancer by using machine learning to more quickly and accurately analyze images of cells.

Its overhead cameras track individuals and items continuously (notably, its so-called entity cohesion doesn’t rely on facial recognition, which it says gives shoppers more privacy).

A lineup of cloud-connected security cameras equipped with AI-driven features like object and movement detection has driven growth at Verkada, whose cofounders are three Stanford computer science graduates and the cofounder of enterprise cloud company Meraki, which sold to Cisco for more than $1 billion.

Among the company’s wide-ranging list of clients are fitness club Equinox, the $1.1 billion Vancouver Mall and more than 500 school districts, which use the cameras for anything from monitoring student safety to tracking food deliveries.

SentinelOne CEO Tomer Weingarten says he and his cofounder started the company in 2013 because antivirus software at the time was “some flavor of bad, incomplete, ineffective, and /or painful to deploy and operate.” They spent the past six years figuring out how to make endpoint security (which focuses on data coming from laptops, phones, and other network-connected devices) smarter, training machine learning models to detect malware in files and running in applications.

“Until now, the most complex operations in manufacturing have been too difficult for blind and dumb robots to perform with the same precision and fidelity as humans,” says CEO Amar Hanspal, adding that advances in computer vision and machine learning have changed the game.

Upstart CEO Dave Girouard admits that most of the early team of former Google employees had no history in financial services when they came up with the idea to apply advanced data science to the credit process in 2012: Only the belief that the current system was antiquated and exclusionary.

By using data not typically found in a person’s credit history to find more nuanced risk patterns, Girouard says Upstart’s lending model has higher approval rates and lower interest rates than traditional methods, with loss rates that are “less than half” of those of peer platforms.

Former Oculus cofounder Palmer Luckey is back after his dramatic exit from Facebook (he has hinted that the company fired him from the virtual reality unit for his political views, which it denies) with a defense technology startup called Anduril Industries, founded in 2017.

It has contracts with the Marine Corps and UK’s Royal Navy, as well as with Customs and Border Protection for what has been described as a controversial “virtual border wall.” Following a report that it became a unicorn after a recent fundraise, the company confirmed to Forbes that it now has $180 million in total funding and a near-billion valuation.

Alexandr Wang’s data labeling startup Scale AI has gained so much attention from customers — particularly autonomous transportation companies, which need gobs of well-labeled data to train their systems — that he’s running a unicorn company before his 23rd birthday.

It pulls public data about a property to automatically answer many of the questions a typical insurer would ask, which means it can quickly dole out quotes, and pulls data from aerial images and smart home sensors to detect issues that could lead to claims in real-time.

Dataminr ingests public internet data, like social media posts, and uses deep learning, natural language processing, and advanced statistical modeling to send users tailored alerts.

Uptake CEO Brad Keywell says his company is in the business of making sure things work, “whether it’s the U.S. Army’s Bradley Fighting Vehicle, or the components that make up Rolls-Royce’s fleet of market-leading engines.” It’s brought in more than 100 industrial customers on its way to a $2.3 billion valuation.

trifecta of autonomy and transportation experts from Tesla, Uber, and Google came together to build Aurora, a self-driving car company that plans to sell its system to automakers instead of operating its own fleet (it currently has a deal with Hyundai to provide software for its future Kia models).

CMNS 253W - On Artificial Intelligence

Spring 2017 - SFU CMNS 253W Prof. Frederik Lesage TA Douglas Amundaray "On Artificial Intelligence" was created to critique and research aspects of ...

youBot Matlab Simulator

youbot sim.

MODERN LAB 6MIN HD

MODERN LAB 6MIN HD.

Spy Project Presentation L.I.N.K. Robot

L.I.N.K. robot follows you and replies to what you say!

Manual Guidance with KUKA youBot (Torque control)

The PD + Gravity compensation control of the joints torque allow to manually move the KUKA youBot arm. Programming by demonstartion is then possible.

HealthBeautyShop.com Mini Cordless Flat Iron! - Dance Version

For more info, please visit The Mini Cordless Flat Iron is a digital cordless rechargeable straightner with a negative ion ..