Artificial synapse on a chip will help mobile devices learn like the human brain

Why it matters to you

Neural networks are responsible for some of the biggest breakthroughs in AI, but they also require a lot of power to run. This synapse on a chip could help.

Brain-inspired deep learning neural networks have been behind many of the biggest breakthroughs in artificial intelligence seen over the past 10 years.

But a new research project from the National Center for Scientific Research (CNRS), the University of Bordeaux, and Norwegian information technology company Evry could take that these breakthroughs to next level — thanks to the creation of an artificial synapse on a chip.

“There are many breakthroughs from software companies that use algorithms based on artificial neural networks for pattern recognition,” Dr. Vincent Garcia, a CNRS research scientist who worked on the project, told Digital Trends. “However, as these algorithms are simulated on standard processors they require a lot of power. Developing artificial neural networks directly on a chip would make this kind of tasks available to everyone, and much more power efficient.”

Synapses in the brain function as the connections between neurons. Learning takes place when these connections are reinforced, and improved when synapses are stimulated. The newly developed electronic devices (called “memristors”) emulate the behavior of these synapses, by way of a variable resistance that depends on the history of electronic excitations they receive.

More: MIT researchers are working to create neural networks that are no longer black boxes

“Here we use a specific kind of memristors based on purely electronic effects,” Garcia continued. “In these devices, the active part is a ferroelectric film, which contains electric dipoles, that can be switched with an electric field. Depending on the orientation of these dipoles, the resistance is on or off. In addition, we can control configurations in which domains with up or down dipoles coexist, with intermediate voltage pulses. This gives rise to an analog device with many resistance levels. In our paper, we were able to understand how the resistance of the memristor evolves with voltage pulses and make a model based on the dynamics of ferroelectric domains.”

The result was an array of 45 such memristors which were able to learn to detect simple patterns without any assistance; something referred to as “unsupervised learning” in the machine learning community.

Now that the team is able to predict the behavior of an individual electronic synapse, the next goal is to develop neural networks on a chip that contain hundreds of these ferroelectric memristors. Once this is achieved, they will test it by connecting the neural network to an event-based camera to have a go at detecting moving objects at high-speed.

“The final goal of this project would be to integrate this bio-inspired camera in a car to assist the driver when unexpected objects or persons are crossing the road,” Garcia said.

Should all go according to plan, it may not be long before the neural networks are incorporated as a standard part of the processors found in our smartphones and other mobile devices.

Artificial synapse on a chip will help mobile devices learn like the human brain

Why it matters to you

Neural networks are responsible for some of the biggest breakthroughs in AI, but they also require a lot of power to run. This synapse on a chip could help.

Brain-inspired deep learning neural networks have been behind many of the biggest breakthroughs in artificial intelligence seen over the past 10 years.

But a new research project from the National Center for Scientific Research (CNRS), the University of Bordeaux, and Norwegian information technology company Evry could take that these breakthroughs to next level — thanks to the creation of an artificial synapse on a chip.

“There are many breakthroughs from software companies that use algorithms based on artificial neural networks for pattern recognition,” Dr. Vincent Garcia, a CNRS research scientist who worked on the project, told Digital Trends. “However, as these algorithms are simulated on standard processors they require a lot of power. Developing artificial neural networks directly on a chip would make this kind of tasks available to everyone, and much more power efficient.”

Synapses in the brain function as the connections between neurons. Learning takes place when these connections are reinforced, and improved when synapses are stimulated. The newly developed electronic devices (called “memristors”) emulate the behavior of these synapses, by way of a variable resistance that depends on the history of electronic excitations they receive.

More: MIT researchers are working to create neural networks that are no longer black boxes

“Here we use a specific kind of memristors based on purely electronic effects,” Garcia continued. “In these devices, the active part is a ferroelectric film, which contains electric dipoles, that can be switched with an electric field. Depending on the orientation of these dipoles, the resistance is on or off. In addition, we can control configurations in which domains with up or down dipoles coexist, with intermediate voltage pulses. This gives rise to an analog device with many resistance levels. In our paper, we were able to understand how the resistance of the memristor evolves with voltage pulses and make a model based on the dynamics of ferroelectric domains.”

The result was an array of 45 such memristors which were able to learn to detect simple patterns without any assistance; something referred to as “unsupervised learning” in the machine learning community.

Now that the team is able to predict the behavior of an individual electronic synapse, the next goal is to develop neural networks on a chip that contain hundreds of these ferroelectric memristors. Once this is achieved, they will test it by connecting the neural network to an event-based camera to have a go at detecting moving objects at high-speed.

“The final goal of this project would be to integrate this bio-inspired camera in a car to assist the driver when unexpected objects or persons are crossing the road,” Garcia said.

Should all go according to plan, it may not be long before the neural networks are incorporated as a standard part of the processors found in our smartphones and other mobile devices.

OnePlus 4 smartphone: News and rumors

Why it matters to you

OnePlus has a history of launching stellar phones at reasonable prices, and it seems likely that its latest offering will be no different.

It seems like it was only yesterday that OnePlus, the controversial Chinese smartphone maker with a cult following, released the latest flagship in its eponymous series: The OnePlus 3T. But rumor has it that a brand-new model is on the horizon. It might be called the OnePlus 4. Here’s everything we know about it so far.

Name

OnePlus might skip the “4” for “5.” India Today reports that the next OnePlus flagship will be called the OnePlus 5 rather than the OnePlus 4, because the number 4 is considered unlucky in Chinese culture. There is precedent — the company chose to label its fourth flagship, the OnePlus 3T, instead of OnePlus 4.  Superstition might have been the reason.

Design

The OnePlus 4’s design remains a mystery, but reports suggest it will adopt an all-glass, curved aesthetic in the style of the Samsungs Galaxy S8’s dual-edge panel.

Despite the rumored glass design, GizBot reports that the OnePlus 4 will be more durable than its predecessor. It’s said to be IP68 rated against water and dust exposure, or safe in up to 1.5m of water for 30 minutes.

More:OnePlus 3T review

The OnePlus 4 will reportedly launch in five different color configurations, including white, black, gold, blue, and ceramic. And another might be on the way: Android Police reporter David Ruddock teased a OnePlus announcement having to do with “Black” and “Midnight,” which might refer to a new midnight black option.

.@getpeid I hear OnePlus could have a dark announcement some time soon. That it might even be as Black as Midnight?

— David Ruddock (@RDR0b11) March 6, 2017

Adobe could start helping you take better selfies via machine learning

Why it matters to you

Appealing to the inner narcissist in all of us, Adobe’s newly demonstrated technology could help us achieve the perfect head tilt for that selfie.

Who needs a photo director when you’ve got yourself, your phone, and Adobe? As per a newly released demonstration of the latest tech slated to emerge from the company, it looks like you could soon be on your way to a stronger selfie game. And who doesn’t need tips and tricks on ideal angles and perfect lighting?

While the iOS app doesn’t have a name yet (and is likely still a ways away from becoming publicly available), it does take advantage of a number of Adobe’s machine learning capabilities. For example, you’ll be able to search for photos that you’d like to emulate with your own selfies (or mobile portraits), and then Adobe will help you apply that style to your photograph. In addition to this feature, Adobe’s new app also seems to be able to turn your two-dimensional selfies into 3D models, smooth lines with the tool’s liquify filter, or even make it seem as though the photo was taken from a different distance.

More: Tired of boring Facebook profile photos? Upload a profile video instead — here’s how

And if you’re a real camera pro, you can edit your iPhone photos to look more like those taken using DSLR cameras by blurring a photograph’s background (paying homage to a depth-of-field blur).

While not all of these features are entirely novel, this will be the first time that they’ve all been presented together and in a mobile app. Previously, many of these tools lived on desktop only. That said, don’t get too excited about being able to use this app too soon. We still don’t have a clear sense of timing, and Adobe has remained evasive about its plans.

“Sneaking glimpses of new technology is something we do from time to time to showcase the amazing innovation that Adobe engineers come up with,” the company in an official statement. “We’ll be sure to keep you updated on future developments of Creative Cloud products as they’re available.”

Future contact lenses may measure glucose, detect cancer, monitor drug use

Why it matters to you

Contact lenses embedded with biosensors may help patients monitor their conditions and even detect diseases early on.

In the future, your contacts lenses may do more than help you see clearly — they may help save your life.

Engineers at Oregon State University are working to embed transparent sensors into contact lenses, which could then monitor biomarkers to track drug use, blood glucose levels, or even detect early signs of cancer.

“My group had been working on a different technology for glucose sensing … which is similar to what is used for glucose test strips,” Greg Herman, a chemical engineer who is leading the research, told Digital Trends. “On a separate project, we were working on transistors that can be transparent. It came to me that we could modify the transistor to be a sensor and make it fully transparent.” The step from conception to creation came quickly, he added.

More: CooperVision’s latest contact lens may be the cure for your digital eye strain

In order to design the biosensors, Herman referred to indium gallium zinc oxide (IGZO), a compound he invented with some colleagues years ago while working in industry. IGZO has already revolutionized touchscreen sensitivity.

Instead of offering consumers higher resolution devices, Herman’s new use for IGZO enables continuous glucose monitoring by carrying an enzyme that reacts with glucose to change the lenses conductivity and signal a patient’s glucose level.

“The thought with the contact lens is that it will give diabetic patients more flexibility in managing their glucose levels,” he said.

Over 2,500 different biosensors could theoretically be embedded in a 1mm square of these contact lenses, according to Herman, detecting things like uric acid levels and cancer biomarkers. However, such a device is still some years off.

“We are still in the early stages, and it is difficult to say how long it will be before they are available commercially,” he said. “We have talked to a couple companies, and are hoping to work with them to accelerate the technology.”

Herman presented his research this week at a meeting of the American Chemical Society.

The HTC Vive isn’t on a diet but it lost weight since its pre-retail days

Why it matters to you

While early adopters of the HTC Vive may feel a financial stab in the gut, the changes HTC made to the headset provides a lighter experience for new owners.

Let’s get real here: The HTC Vive virtual reality headset isn’t exactly like wearing sunglasses. It’s big and bulky, not to mention it drags a really long cord across the floor. If anything, the HTC Vive could roughly have the same weight as a decent-sized ceramic bowl. But the Vive was heavier at launch than it is now, so there is hope that the head-mounted experience will get even lighter in the next generation.

When Vive arrived in April 2016, it weighed 555 grams (1.2 pounds). But after a year of refining the overall design, HTC’s most recent Vive headset weighs around 470 grams (1 pound). That is a reduction of 85 grams just in the headset alone, which equals to three ounces. While that does not seem like much of a reduction, your face will likely disagree.

More: HTC Vive users can start playing without wires starting at the end of April

“We are always looking to improve the core Vive experience in terms of comfort and durability,” HTC’s Patrick Seybold told Digital Trends. “We’ve made several improvements since launch including a new 3-in-1 cable. We’ve also made smaller improvements to the weight due to component swaps and improvements in manufacturing. This has resulted in a 15 percent decrease since the Vive Pre.”

The Vive Pre, of course, was the near-final, pre-retail version of HTC’s VR headset. The updated 3-in-1 cable Seybold refers to seemingly made its appearance in August and is now part of the current HTC Vive package. However, owners of the older models can take the “upgrade” plunge by replacing the previous cable for the new $40 version right here. It measures 16.4 feet long, so continue to watch your step.

That said, with the combined weight reduction in the headset and the refined cable, the latest Vive retail package sold on the market today is significantly lighter than the one sold at launch. Product refinement and weight loss should only increase as the headset moves into toddler territory.

Another improvement to the HTC Vive experience will be the upcoming Vive Deluxe Audio Strap, which retails for $100 this June. It replaces the current strap packed in the retail box while eliminating the need for the annoying ear buds. There is nothing like jerking your head left or right and the earbuds suddenly rip out, disrupting the immersive experience. The “deluxe” strap providing built-in headphones will eliminate that audio disruption.

HTC’s “deluxe” strap will also throw out all the Velcro in favor of an adjustment dial to easily meet the needs of every skull size. Even more, users will have the option of placing the cable path off to the side rather than across the top of the user’s head. The built-in headphones are capable of vertical adjustments too.

Vive Day took place on Wednesday, celebrating the headset’s first birthday. As part of the celebration, Steam is currently offering a load of discounted Vive-compatible games and experiences until Monday. These include the Serious Sam VR bundle ($69), Arizona Sunshine, ($34), Raw Data ($32), and more.

Lenovo shows off the Blade, another novel Windows 10 2-in-1 concept

Why it matters to you

With concepts like Lenovo’s latest Blade, there should be no shortage of innovative new devices joining the Windows 10 PC ecosystem.

If you’ve been following the Windows 10 PC ecosystem over the last few years, then you’ve no doubt noticed that there’s no shortage of innovative form factors. The operating system’s focus on touch and pen input, not to mention Microsoft’s own Surface devices, have compelled PC makers to come up with some fascinating designs.

Lenovo, in particular, hasn’t been shy about coming up with something new. Its Yoga Book 2-in-1 device is a case in point, and now the company is pushing the envelope again with its Blade “next generation laptop” that offers up its own innovations, as Liliputing points out.

More: Lenovo Yoga Book Review

The Lenovo Blade garnered some attention at the most recent IF World Design Guide Awards, where Lenovo managed to score a number of wins. The Blade is a Windows 10 2-in-1 in a tablet format utilizing a detachable keyboard and a built-in cover that folds over to protect the screen when not in use. In that respect, it’s not unlike a number of other 2-in-1 machines.

Where the Blade diverges from the pack is in the use of Miracast wireless display technology, which allow the keyboard and display portion to continue working together even when the keyboard is detached. A magnetic connector is used to easily attached and detach the keyboard. But when you pull them apart, the machine keeps working, enabling some serious flexibility.

In addition, the Blade can connect with other Miracast devices, meaning that its functionality could be expanded in the future. It’s not clear from the information that Lenovo has released so far exactly how the system works, and whether the computing technology is in the keyboard, the display, or both.

Regardless, the Lenovo Blade is a fascinating concept that would introduce yet another unique form factor to the Windows 10 PC ecosystem. There’s no word yet on if or when Lenovo might release the Blade to the market, or how it will be priced.

BioLite offers new ways to keep our gadgets charged in the backcountry

Why it matters to you

BioLite delivers clean, efficient, and portable power to the backcountry with its new line of USB chargers and an upgraded solar panel.

Keeping our mobile devices charged in the backcountry just got a little easier. BioLite, the company behind the battery-charging CampStove, has announced a slew of new products specifically designed to power our smartphones, tablets, headlamps, cameras, and other gadgets while traveling off the grid, including a new portable solar panel and a line of USB battery packs.

Last year, BioLite introduced its first two solar panels, bringing a lightweight and efficient solution for charging electronic devices to the campsite. Those models included the SolarPanel 5 and the SolarPanel 5+, each of which are capable of generating up to 5 watts of power from the sun. The difference between the two models is that the “+” version includes a built-in battery pack with a capacity of 2,200mAh to store the energy collected. Now, the company is expanding the line to include a third option.

The new SolarPanel 10+ also comes equipped with a 3,000mAh battery pack and is capable of generating as much as 10 watts of power thanks to an additional fold-out solar cell. Like BioLite’s previous models, it has a built-in kickstand to keep it stable on a variety of surfaces, as well as a sundial to help users find the most efficient angle and positioning to collect power. The SolarPanel 10+ sells for $130 and is available now.

The SolarPanel 10+ isn’t the only new addition to the BioLite catalog. The company has also taken the wraps off its new Charge line of USB battery packs as well. Featuring a thin and lightweight — yet durable — stainless steel body, these portable chargers are built to be tough enough to take with you on your outdoor adventures, but elegant enough for use in your day-to-day life, too. The Charge 10 includes a single USB port and a 2,600mAh battery, and sells for $25. The Charge 20 offers two USB ports and a 5,200mAh battery, and is available at a $40 price point. Both will begin shipping on April 15, while a third addition to the Charge line — the Charge 40 — will ship later this summer. It is expected to have a 10,400mAh battery.

More: The Klymit Double V is an inflatable sleeping pad made for camping couples

BioLite says that all of the Charge models are “dunkproof,” which essentially means they can survive getting splashed with water or exposed to some rain. But they are not fully waterproof so you should expect them to survive getting dropped in a river or lake. That said, they should be durable enough to survive most of the other abuse they’ll receive in the outdoors.

Robotic whiteboard draws tweets, messages, and artwork on command

Why it matters to you

With Joto you can satisfy your nostalgia for written text without taking a step back in time.

Have you ever wanted to send a distant loved one a handwritten message but didn’t have the patience for post? A design studio called Those has you covered.

Available via Kickstarter, Joto is a connected display board that draws tweets, messages, and even art with a pen. The device combines digital connectivity with analog aesthetic. It integrates with services such as Twitter, Trello, and Slack to manifest tweets, tasks, and messages on command.

More: Whip out your Apple Pencil and try out the 20 best iPad Pro drawing apps

Joto is the actually the successor of a lightweight drawing machine called Woodpecker, intended to let retail stores and their customers scribble on windows.

“I used to work in a retail department store and was frustrated by how little beyond printed graphics were being used to engage customers,” Those founder Jim Rhodes told Digital Trends. “That’s when where the idea for an interactive window display came from. With the arrival of 3D printing and the maker movement a few years later, we were able to make it a reality.”

Where Woodpecker was designed mainly to engage customers, Joto is intended to fit more intimately into homes and workspaces. Want to tell your partner you’ll be home late? Ask Joto to deliver the message. Need to remind your colleague about that upcoming presentation? Send them a drawing of a ticking clock. But if you’re all out of ideas, you can ask Joto to improvise.

“Joto was designed to take requests,” Rhodes said. “We haven’t planned to make Joto intelligent just yet, but is part of a new movement away from the screen. Not because screens are bad but because people are discovering new ways to interact with internet.”

You can back Joto on Kickstarer for $199. Add an extra $100 for a year’s worth of daily curated illustrations. As of publishing this article, it has raised 256,335 British pounds ($320,034) with a 100,000-pound ($124,850) goal and six days left in the campaign. Those hopes to deliver Jotos to Kickstarter backers by December. The product’s official launch is set for March 2018.

The 8 most common problems with Windows 10 Creators Update, and how to fix them

The Windows 10 Creators Update is finally here, if you’re brave enough to try grabbing the Windows Insider Preview that is. As a major update to Windows, the Creators Update offers a host of new features designed to maximize your creativity and help you get the most out of your operating system. Alongside marquee features like Paint 3D and Game Mode, the Creators Update introduces dozens of smaller ease-of-use improvements, and tons of minor changes throughout Windows 10.

Like every major OS update, however, the Creators Update is not without its problems. Users have flocked to the Windows support forums and Windows Insider subreddits to express their frustrations and describe in detail how this latest OS update is misbehaving. So, we’ve put together a list of the eight most common Windows 10 Creators Update problems, and how you can fix them.

More: Official Windows 10 Creators Update disk image now available for Insiders

The update gets stuck

Sometimes the hardest part of getting a new Windows update to work is getting the update itself. If you’ve found that your system keeps trying to download and install the Creators Update but fails partway through or just doesn’t apply the update no matter how many times you hit “Restart and update,” you might need to fire up your command prompt.

Windows Update doesn’t always do its job well, so if you’re experiencing problems installing the Creators Update first you should try clearing out the existing Update files. You can do that by opening up your command prompt, just type “CMD” into your Windows Search and it should pop up — make sure you right click it and hit “Run as administrator” or these fixes might not work.

First up, when the command prompt opens, type “net stop wuauserv” and hit enter. This will temporarily disable your Windows Update. Next, navigate to C:WindowsSoftwareDistribution and delete the contents of this folder but not the folder itself. These are the files Windows Update is trying to use to update your Windows install.

Finally, go back to your command prompt and type “net start wuauserv” and hit enter. This will restart the Windows Update service. Now, fire up Windows Update, and re-download the Creators Update. It  should work now.