Connect with us


Non-invasive glucose monitor EasyGlucose takes home Microsoft’s Imagine Cup and $100K – TechCrunch



Microsoft’s yearly Imagine Cup student startup competition crowned its latest winner today: EasyGlucose, a non-invasive, smartphone-based method for diabetics to test their blood glucose. It and the two other similarly beneficial finalists presented today at Microsoft’s Build developer conference.

The Imagine Cup brings together winners of many local student competitions around the world, with a focus on social good and, of course, Microsoft services like Azure. Last year’s winner was a smart prosthetic forearm that uses a camera in the palm to identify the object it is meant to grasp. (They were on hand today as well, with an improved prototype.)

The three finalists hailed from the U.K., India and the U.S.; EasyGlucose was a one-person team from my alma mater UCLA.

EasyGlucose takes advantage of machine learning’s knack for spotting the signal in noisy data, in this case the tiny details of the eye’s iris. It turns out, as creator Bryan Chiang explained in his presentation, that the iris’s “ridges, crypts and furrows” hide tiny hints as to their owner’s blood glucose levels.

EasyGlucose presents at the Imagine Cup finals

These features aren’t the kind of thing you can see with the naked eye (or rather, on the naked eye), but by clipping a macro lens onto a smartphone camera, Chiang was able to get a clear enough image that his computer vision algorithms were able to analyze them.

The resulting blood glucose measurement is significantly better than any non-invasive measure and more than good enough to serve in place of the most common method used by diabetics: stabbing themselves with a needle every couple of hours. Currently EasyGlucose gets within 7% of the pinprick method, well above what’s needed for “clinical accuracy,” and Chiang is working on closing that gap. No doubt this innovation will be welcomed warmly by the community, as well as the low cost: $10 for the lens adapter, and $20 per month for continued support via the app.

It’s not a home run, or not just yet: Naturally, a technology like this can’t go straight from the lab (or in this case, the dorm) to global deployment. It needs FDA approval first, though it likely won’t have as protracted a review period as, say, a new cancer treatment or surgical device. In the meantime, EasyGlucose has a patent pending, so no one can eat its lunch while it navigates the red tape.

As the winner, Chiang gets $100,000, plus $50,000 in Azure credit, plus the coveted one-on-one mentoring session with Microsoft CEO Satya Nadella.

The other two Imagine Cup finalists also used computer vision (among other things) in service of social good.

Caeli is taking on the issue of air pollution by producing custom high-performance air filter masks intended for people with chronic respiratory conditions who have to live in polluted areas. This is a serious problem in many places that cheap or off-the-shelf filters can’t really solve.

It uses your phone’s front-facing camera to scan your face and pick the mask shape that makes the best seal against your face. What’s the point of a high-tech filter if the unwanted particles just creep in the sides?

Part of the mask is a custom-designed compact nebulizer for anyone who needs medication delivered in mist form, for example someone with asthma. The medicine is delivered automatically according to the dosage and schedule set in the app — which also tracks pollution levels in the area so the user can avoid hot zones.

Finderr is an interesting solution to the problem of visually impaired people being unable to find items they’ve left around their home. By using a custom camera and computer vision algorithm, the service watches the home and tracks the placement of everyday items: keys, bags, groceries and so on. Just don’t lose your phone, as you’ll need that to find the other stuff.

You call up the app and tell it (by speaking) what you’re looking for, then the phone’s camera determines your location relative to the item you’re looking for, giving you audio feedback that guides you to it in a sort of “getting warmer” style, and a big visual indicator for those who can see it.

After their presentations, I asked the creators a few questions about upcoming challenges, since as is usual in the Imagine Cup, these companies are extremely early-stage.

Right now EasyGlucose is working well, but Chiang emphasized that the model still needs lots more data and testing across multiple demographics. It’s trained on 15,000 eye images but many more will be necessary to get the kind of data they’ll need to present to the FDA.

Finderr recognizes all the images in the widely used ImageNet database, but the team’s Ferdinand Loesch pointed out that others can be added very easily with 100 images to train with. As for the upfront cost, the U.K. offers a £500 grant to visually-impaired people for this sort of thing, and they engineered the 360-degree ceiling-mounted camera to minimize the number needed to cover the home.

Caeli noted that the nebulizer, which really is a medical device in its own right, is capable of being sold and promoted on its own, perhaps licensed to medical device manufacturers. There are other smart masks coming out, but he had a pretty low opinion of them (not strange in a competitor, but there isn’t some big market leader they need to dethrone). He also pointed out that in the target market of India (from which they plan to expand later) it isn’t as difficult to get insurance to cover this kind of thing.

While these are early-stage companies, they aren’t hobbies — though, admittedly, many of their founders are working on them between classes. I wouldn’t be surprised to hear more about them and others from Imagine Cup pulling in funding and hiring in the next year.

Source link

Continue Reading
Click to comment

Leave a Reply

Your email address will not be published. Required fields are marked *


One of the Internet’s most aggressive threats could take UEFI malware mainstream



One of the Internet’s most aggressive threats has just gotten meaner, with the ability to infect one of the most critical parts of any modern-day computer.

Trickbot is a piece of malware that’s notable for its advanced capabilities. Its modular framework excels at gaining powerful administrator privileges, spreading rapidly from computer to computer in networks, and performing reconnaissance that identifies infected computers belonging to high-value targets. It often uses readily available software like Mimikatz or exploits like EternalBlue stolen from the National Security Agency.

Once a simple banking fraud trojan, Trickbot over the years has evolved into a full-featured malware-as-a-service platform. Trickbot operators sell access to their vast number of infected machines to other criminals, who use the botnet to spread bank trojans, ransomware, and a host of other malicious software. Rather than having to go through the hassle of ensnaring victims themselves, customers have a ready-made group of computers that will run their crimeware.

The first link in the security chain

Now, Trickbot has acquired a new power: the ability to modify a computer’s UEFI. Short for Unified Extensible Firmware Interface, UEFI is the software that bridges a computer’s device firmware with its operating system. As the first piece of software to run when virtually any modern machine is turned on, it’s the first link in the security chain. Because the UEFI resides in a flash chip on the motherboard, infections are difficult to detect and remove.

According to research findings published on Thursday, Trickbot has been updated to incorporate an obfuscated driver for RWEverything, an off-the-shelf tool that people use to write firmware to virtually any device.

At the moment, researchers have detected Trickbot using the tool only to test whether an infected machine is protected against unauthorized changes to the UEFI. But with a single line of code, the malware could be modified to infect or completely erase the critical piece of firmware.

“This activity sets the stage for TrickBot operators to perform more active measures such as the installation of firmware implants and backdoors or the destruction (bricking) of a targeted device,” Thursday’s post jointly published by security firms AdvIntel and Eclypsium stated. “It is quite possible that threat actors are already exploiting these vulnerabilities against high-value targets.”

Rare for now

So far, there have been only two documented cases of real-world malware infecting the UEFI. The first one, discovered two years ago by security provider ESET, was done by Fancy Bear, one of the world’s most advanced hacker groups and an arm of the Russian government. By repurposing a legitimate antitheft tool known as LoJack, the hackers were able to modify UEFI firmware so that it reported to Fancy Bear servers rather than ones belonging to LoJack.

The second batch of real-world UEFI infections were uncovered only two months ago by Moscow-based security firm Kaspersky Lab. Company researchers found the malicious firmware on two computers, both of which belonged to diplomatic figures located in Asia. The infections planted a malicious file in a computer’s startup folder so it would run whenever the computer booted up.

The motherboard-resident flash chips that store the UEFI have access control mechanisms that can be locked during the boot process to prevent unauthorized firmware changes. Often, however, these protections are turned off, misconfigured, or hampered by vulnerabilities.

UEFI infections at scale

At the moment, the researchers have seen Trickbot using it’s newly acquired UEFI-writing capabilities to test if the protections are in place. The presumption is that the malware operators are compiling a list of machines that are vulnerable to such attacks. The operators could then sell access to those machines. Customers pushing ransomware could use the list to overwrite the UEFI to make large numbers of machines unbootable. Trickbot clients intent on espionage could use the list to plant hard-to-detect backdoors on PCs in high-value networks.

Trickbot’s embrace of UEFI-writing code threatens to make such attacks mainstream. Instead of being the dominion of advanced persistent threat groups that typically are funded by nation states, access to UEFI-vulnerable computers could be rented out to the same lower-echelon criminals who now use Trickbot for other types of malware attacks.

“The difference here is that TrickBot’s modular automated approach, robust infrastructure, and rapid mass-deployment capabilities bring a new level of scale to this trend,” AdvIntel and Eclypsium researchers wrote. “All pieces are now in place for mass-scale destructive or espionage-focused campaigns that can target entire verticals or portions of critical infrastructure.”

Continue Reading


Apple’s MagSafe Duo charger finally shows up in online stores



Today, Apple finally began selling and shipping the MagSafe Duo charger, an accessory that was announced alongside the new iPhone 12 lineup on October 13.

The MagSafe Duo is a charging pad that uses Qi wireless charging tech to charge two devices at once. Apple says it supports any Qi-enabled device but, of course, it specifically notes that the Duo can charge any two-device combination of an iPhone, an Apple Watch, or AirPods cases.

The MagSafe Duo charger is just one of a number of MagSafe accessories Apple announced in October. New iPhones have a magnet built into the back that is designed to allow chargers and cases to latch on, and MagSafe iPhones also have built-in sensors to detect what kind of accessory is being connected and communicate with it via NFC.

Apple also announced at that same iPhone event that new iPhones wouldn’t come with charging bricks in their boxes, and that also applies to this charger. Apple is assuming here that most anyone who buys this already has a charging brick on hand, but if you don’t, Apple’s store page says the following about what you should make sure to get to use the MagSafe Duo:

– 20W USB-C Power Adapter (sold separately) for faster wireless charging up to 11W
– 27W or higher USB-C Power Adapter (sold separately) for faster wireless charging up to 14W

Obviously, wired charging is going to be much faster. But if you like the convenience of wireless charging, this is Apple’s intended multidevice solution for you—though it doesn’t support three devices at once like the long-ago announced-then-seemingly-canceled AirPower mat.

The MagSafe Duo is priced at $129 and ships right away, with the first deliveries expected to arrive at buyers’ doorsteps later this week.

Listing image by Apple

Continue Reading


Qualcomm’s new flagship SoC is the Snapdragon 888



Enlarge / The Snapdragon 888, sitting on the world’s biggest ARM motherboard.


This week Qualcomm announced its flagship smartphone SoC for 2021, the “Snapdragon 888.” The TL;DR is that Qualcomm’s 2021 chip is a 5nm SoC with an ARM Cortex-X1 core and Qualcomm’s first flagship SoC with an on-board 5G modem, dumping the mandatory two-chip 5G solution that Qualcomm forced on the industry earlier this year with the Snapdragon 865. Compared to the Snapdragon 865, Qualcomm is promising performance improvements of 25 percent from the CPU, 35 percent from the GPU, and 35 percent from the ISP.

We should probably talk about the name first. Qualcomm’s normal naming scheme (and the rumor mill) would have made this chip the “Snapdragon 875” after the 865, 855, and 845 from previous years. The switch to Snapdragon 888 is apparently a nod to Chinese culture, which views 8 as a lucky number. Flight numbers out of Chinese airports often have a lot of eights in them, the Beijing Olympics began on 8/8/08 at eight seconds past 8:08pm local time, and now flagship Android phones will somehow be more appealing—I guess—to Chinese consumers, which also happen to make up the biggest smartphone market in the world. Marketing!

As usual, the CPU is sporting a very lucky eight cores with a single “Prime” core for higher performance duties, three medium cores to help out with foreground tasks, and four low-power cores for background processing. This year the prime core is getting upgraded to ARM’s big, new Cortex-X1 core at 2.84GHz, while the medium cores are getting upgraded to the Cortex A78. The ancient A55 core is still working smaller-core duties.

ARM’s Cortex A78 core is the successor to the Cortex A76 that was used in the previous-gen Snapdragon 865, and ARM claims the design—including the drop from 7nm to 5nm—results in 20-percent better “sustained performance” in the same thermal envelope as a 7nm Cortex A76. The A78 was designed with ARM’s typical focus on “performance, power, and area (PPA),” meaning it strikes a balance between physical size, power usage, and CPU performance. The new big core this year is the Cortex-X1, which was designed alongside the A78 as a whole new size category for ARM. The X1 is a super-sized A78 that asks, “What if we don’t worry about power usage and size?” and just go all-out for performance, and the result is a core that ARM says adds “30 percent peak performance” over the A76. The idea is that the A78 is here for sustained performance, while the X1 is here to increase the cap on bursty CPU performance, which is useful for loading apps and webpages.

Qualcomm never goes into much detail about its GPUs, but it does support Variable Rate Shading (VRS) this year, which lets game developers finely control how much detail each part of the display is being rendered in from frame to frame. If a scene is dark, or an object is moving very quickly, it probably doesn’t need to be rendered at full resolution, allowing clever developers to squeeze more power out of a GPU by turning down only the parts you won’t notice. VRS debuted on PCs with Nvidia’s Turing GPUs in 2018, and it’s supported in Vulkan and Unreal Engine (and DirectX), so there’s a shot some mobile games will actually use it.

The block diagram shows Qualcomm’s “FastConnect 6900 System” is integrated on the 888 SoC, which is Qualcomm-speak for the Wi-Fi and Bluetooth connectivity. This means phones will be able to support Bluetooth 5.2 and the new Wi-Fi 6E, if the manufacturers put in the extra work and components to take advantage of it. Wi-Fi 6E will eventually be a big deal for people in dense living areas, since it takes all the advantages of Wi-Fi 6 and adds a big chunk of 6GHz spectrum to the standard. Every country’s spectrum regulations are different, but in the US, Wi-Fi 6E will have access to three times the total spectrum of current Wi-Fi systems, or six times the useful spectrum if you skip counting the DFS allocated spectrum in the middle of the 5GHz Wi-Fi bands, which most access points ignore by default. This means Wi-Fi 6E clients will work a lot better in crowded areas like apartment buildings, where the huge number of close-together access points and clients can fill the airwaves and slow the network to a crawl. Early adopters will get the new spectrum all to themselves in the early days, and as time goes on, even the Wi-Fi laggards will see less competition for the airwaves as everyone else switches over.

Wi-Fi 6E compatibility was first possible for Qualcomm on the Snapdragon 865 Plus, but (while there have been a few false alarms) we’ve yet to see “The World’s First Wi-Fi 6E Phone” hit the market. Unless someone beats Samsung to the punch, it looks like the Galaxy S21 Ultra (not any of the cheaper variants) will be the first to come with Wi-Fi 6E in early 2021. We’ve also yet to see a 6E access point, which you’ll also need to take advantage of the new spectrum. We’ve got to start somewhere, though, and that “somewhere” will be flagship Android phones.

Fixing 5G, and other questionably useful additions

One big improvement of the Snapdragon 888 is undoing the damage of the Snapdragon 865, which split Qualcomm’s SoC into two chips: a main SoC and a separate modem. In order to rush 5G more into the mainstream market, Qualcomm split the modem off into a separate chip, and it brought 4G along for the ride, too. A two-chip solution is generally bigger, hotter, more power-hungry, and more expensive than a one-chip solution. Sure enough, 2020 saw phone prices skyrocket, with bigger bodies and bigger batteries to handle the higher power draw and heat of Qualcomm’s multi-chip 5G solution. While it’s expected that an early-generation 5G modem would come with size and power tradeoffs, Qualcomm moved 4G to the separate modem, too, bringing the power-usage downsides to 4G and making it so you couldn’t just turn off 5G in areas without 5G coverage, which is most areas.

The Snapdragon 888 is a return to sanity, with onboard 4G and 5G connectivity. We’re still not quite ready to say “5G is ready for prime-time” given the sorry state of 5G networks, but it sounds like the hardware is getting there, at least. The Snapdragon 888 is equipped with the Qualcomm X60 modem, which was announced all the way back in February (for some reason Qualcomm modems get announced a year ahead of time). The X60 isn’t a speed increase over 2020’s X55 modem (7.5Gbps down, good luck actually getting that), but it supports all sorts of future-facing network capabilities. There’s Dynamic Spectrum Sharing, which lets carriers run a 5G network over the 4G spectrum. 5G Voice-over-NR (VoNR), aka phone calls over 5G, will be great when we eventually kill 4G LTE. The X60 can also connect to sub-6GHz 5G and mmWave 5G simultaneously, and there’s even support for multiple 5G SIMs.

Other than that, there are a lot of other questionably useful odds and ends packed into Qualcomm’s latest SoC:

  • Qualcomm says the Hexagon 780 co-processor is “completely redesigned” and offers a “three-fold improvement” in performance-per-watt compared to the Snapdragon 865. You’d need an app to use the appropriate AI framework to take advantage of this extra silicon, and most apps don’t, but Qualcomm calls out camera apps, voice assistants, and games as possible targets. Qualcomm also says it will make Snapchat faster.
  • The new ISP supports capturing from three cameras simultaneously. I sort of understood the idea when Qualcomm added two camera streams, so you could record from the front and back camera at the same time, (Nokia obnoxiously called this a “bothie” instead of a selfie) but three? Qualcomm says this would make switching camera lenses faster, but that already happens in a fraction of a second on the Snapdragon 865. I guess the Nokia 9 needed an extra chip to capture data from five rear cameras simultaneously, though that technique didn’t lead to a great end result. Google’s Project Tango needed three cameras for 3D sensing (though today you can do AR with a single camera). Maybe someone will cook up some crazy AI-powered use-case for this.
  • There’s now a Type-1 hypervisor built into the Snapdragon 888, which Qualcomm pitches as “a new way to secure and isolate data between apps and multiple operating systems on the same device.” That sounds very interesting, but I’m not sure what new features it’s supposed to bring to consumer devices. This is “the Android chip” (ARM laptops would use the Snapdragon 8cx line), and isolated apps and data are already more conveniently available in Android via the Work profile, rather than split across operating systems. You can also already boot multiple installs of Android—provided one is a guest OS—for testing. In the past, Google has explicitly shut down attempts at dual-booting Android with other OSes, so don’t let your imagination run too wild. Maybe modders would like it.

Your first shot at owning a Snapdragon 888 will probably be the Samsung Galaxy S21, which is due out in Q1 2021.

Continue Reading