MIT solves a major problem holding up glasses-free 3D TVs

A year ago, MIT scientists unveiled a new technology that delivered a 3D movie experience in theaters without the need for cumbersome glasses. Now they’re working to bring that technology to your living room so that one day you can watch 3D films in your home theater without eyewear.

Their system, “Home3D,” translates typical stereoscopic 3D movies to “multiview,” which shows three or more images to project the scene in three dimensions, portraying it to viewers as if they were really there. MIT’s system converts from cinema-specced stereo to a format compatible with “automultiscopic displays,” which enable the glasses-less experience.

In a paper, the CSAIL team described how their Home3D system uses an algorithm to combine two existing stereo-to-multiview conversion techniques, phase-based and depth image-based rendering, for a superior experience. The algorithms also let users tone down the 3D whenever they wish. While Home3D can currently run on anything with a GPU — an Xbox One or a PlayStation 4, for example — the team believes a future version could fit on a chip to include in smart TVs or add-ons like Chromecast. While glasses-less 3D is still in the early stages where lack of content and lack of hardware support has kept it from gaining momentum, the MIT CSAIL team’s tech could solve half that problem by enabling existing media to play nice with automultiscopic displays.

Apple builds data center to obey Chinese cybersecurity rules

Apple is building its first data center in China in a bid to speed up services for local users and adhere to convoluted cybersecurity laws introduced by the country last month. According to Chinese officials the regulations, which state that all foreign firms must store their data in China, are designed to counter the threat of cyberattacks.

While many overseas businesses claim the law is vague and burdensome, Apple seems more than happy to continue courting its second-largest market. The center, which will be built in the Southern province of Guizhou, is part of a planned $1 billion investment in the province, and comes less than a year after the tech giant announced plans to open a research and development center in Shenzen.

The new facility will be driven entirely by renewable energy and run in partnership with data management firm Guizhou-Cloud Big Data Industry Co Ltd. “The addition of this data center will allow us to improve the speed and reliability of our products and services while also complying with newly passed regulations,” the company said in its statement. “Apple has strong data privacy and security protections in place and no backdoors will be created into any of our systems.”

Via: CNBC

Apple’s Education Store is Down, Back to School Promotion Incoming?

An eagle-eyed Reddit user has noticed that Apple’s higher education store is currently unaccessible in countries including the United States, Canada, Mexico, and Singapore, fueling speculation that Apple could be preparing to launch its annual Back to School promotion for students and educators.

“We’ll be back,” the page reads, which is Apple’s standard message when its website is being updated with new products and information. However, the downtime could also be nothing more than routine maintenance.

For over a decade, Apple has offered an annual Back to School promotion, offering K-12 and college or university students, parents of students, and educators incentives, such as a free pair of Beats headphones or an Apple Store gift card valued up to $100, with the purchase of a qualifying Mac, iPhone, or iPad.

Since 2006, Apple has launched its Back to School promotion in the United States and Canada as early as May 25, and as late as July 23. The promotion is usually extended to several European countries such as France, Germany, and Italy on either the same day, or no more than a few weeks later.

These have been the promotion’s exact kickoff dates each year in North America:

• 2016: June 2
• 2015: July 23
• 2014: July 1
• 2013: July 2
• 2012: June 11
• 2011: June 16
• 2010: May 25
• 2009: May 27
• 2008: June 3
• 2007: June 5
• 2006: June 5

Apple’s Back to School promotion is highly anticipated because it’s one of the few times a year that Apple offers deals to customers, and many hold off on summer purchases until the event begins. Apple offered free Beats headphones in 2016 and 2015, free Apple Store gift cards in 2014, and free iTunes gift cards in 2013.

Apple launched a Back to Uni promotion in Australia and New Zealand in February, offering a free Apple Store gift card valued between $70 and $100 with the purchase of a qualifying new Mac or iPad Pro.

Tags: education, Back to School promotion, Apple retail
Discuss this article in our forums

You Can Now Try HomeKit at Dozens of Apple Stores Around the World

Apple has recently unveiled interactive HomeKit experiences in 46 of its retail stores worldwide, allowing customers to test out its smart home platform free of charge, according to TechCrunch.

Interactive HomeKit demo area at an Apple Store via TechCrunch
Each interactive setup consists of two vertical displays positioned behind an iPhone, iPad, and Apple Watch with the Home app, which customers can use to control accessories in the virtual room, such as lights and ceiling fans.

Now, when you go into Apple’s new retail stores, you’ll be able to use the Home app from either an Apple Watch, iPhone or iPad to control devices like the Phillips Hue light bulb, the Hunter ceiling fan and many others. If you tap to the lower the shades in the living room, for example, you’ll see the shades lower in the house shown on the screen.

In the United States, customers can try HomeKit at Apple’s flagship Union Square store in San Francisco, its World Trade Center and Williamsburg stores in New York City, and 28 other stores not named throughout the country.

The experience is also available until at least December in 15 stores outside of the United States, including select locations in the United Kingdom, Germany, Mexico, Singapore, Taiwan, and the United Arab Emirates.

At every other Apple retail store, the company will be offering non-interactive HomeKit experiences, according to the report.

HomeKit enables Apple users to control lights, switches, thermostats, fans, and other products with its Home app and Siri voice commands. HomeKit-enabled accessories can be controlled singularly, or in scenes, which enable multiple accessories to work in combination, all with a single command.

Tags: HomeKit, Home, Apple retail
Discuss this article in our forums

Samsung’s latest acquisition could help Bixby sound more human

Why it matters to you

Technology that makes virtual assistants more human-sounding, and easier to understand, will make them more useful to us in the future.

Samsung has acquired Innoetics, a company specializing in speech technology, which it may use to improve its Bixby virtual assistant, or use in its rumored smart home speaker system. Innoetics works primarily in text-to-speech, using incredibly natural sounding digital voices, but also develops voices for artificial intelligence, and even digital characters for use in talking books. Its technology has seen the creation of 29 different “synthetic voices,” which can speak 15 different languages.

It’s obvious how Samsung could benefit from Innoetics technology and expertise, but the company hasn’t talked about what its intentions are yet. Samsung has confirmed the acquisition, and said only that Innoetics technologies, “present an opportunity to strengthen Samsung’s capabilities,” in a statement to Techcrunch.

Innoetics, which is based in Greece, hasn’t worked on apps or consumer projects before, and has instead collaborated with local carriers and companies such as Duolingo. Following the acquisition it will end its current services and “focus on further developing the technology it has planned and developed so far,” according to co-founder Emilios Halamandaris, speaking to local newspaper Ekathimerini. The same publication quotes anonymous sources saying the voice technology will be used by Samsung to challenge Apple and Google, further suggesting it will become part of the Bixby virtual assistant.

We can see a hint of Innoetics ambitions in this area from a blog post the company published at the end of 2016, saying it was working on the most advanced synthetic voices it had ever designed, which were ideal for digital assistants due to the human-like tone, and ability to give the assistant a unique character. Apple and Amazon have already achieved this, with recognizable voices and mannerisms for Siri and Alexa. Samsung will want to do the same with Bixby.

Bixby had a difficult start in life, despite Samsung acquiring AI experts — and the firm behind Siri — Viv Labs to help out with development, and the assistant wasn’t ready for international use when the Galaxy S8 launched. It’s still having problems, and is likely to get even more exposure soon. It’s almost certainly going to be a feature on the Galaxy Note 8, a very high-profile release expected in the next few months, and may be part of a future smart home speaker to take on Amazon’s Echo, Google Home, and Apple’s HomePod. Making Bixby sound, read, and interact more naturally is logically a high priority for Samsung. However, it’s unlikely Innoetics technology will ever be directly highlighted by the company, but news of the acquisition gives us an interesting insight into Samsung’s future plans.

AI trained on 1990s blockbuster movies learns to identify handguns on video

Why it matters to you

This artificial intelligence may be used to keep people safe by spotting handguns in places they shouldn’t be.

Humans are hardwired to quickly spot objects in the corner of our eyes: Kids getting into trouble, snakes slithering into the room, and the like. That task isn’t as easy for algorithms, but they’re getting better. Someday soon they may even help humans spot dangerous objects we might have missed. Researchers at the University of Granada, for example, have used deep learning to train an algorithm to identify handguns from video in real time.

The system may be used in home security systems, police body cameras, and video surveillance at places like schools, airports, and malls. It could also be applied to videos online as a way to monitor and map violent content.

Although around 30 percent of American adults say they own guns; the weapons have a steady presence in our lives. They seem to be everywhere you look — carried by cops on TV or as a key component in many of today’s most popular video games. So it’s appropriate that the University of Granada researchers trained their system on movies like Pulp Fiction, and the Mission Impossible and James Bond franchises, through which many of us also had our first introduction to firearms.

After training the algorithm on low-quality video of these classic films, it was able to effectively identify drawn handguns more than 96.5 percent, while analyzing in real time at 5 frames per second. When the algorithm spots a handgun, it borders it with a red box. The researchers think that the model can be combined with an alarm system in order to quickly and inexpensively identify firearms in places where they’re forbidden or may pose a threat.

Although the scientists don’t expect their model to completely replace current weapon-detection systems like metal detectors, they see it as a complementary system to help address some of the drawbacks of conventional methods. For example, metal detectors require guns to pass through a certain point, which — as anyone whose passed a security checkpoint knows — can create a bottleneck of people in crowded areas. The model is also able to discriminate between guns and other metal objects, which the metal detector cannot.

AI trained on 1990s blockbuster movies learns to identify handguns on video

Why it matters to you

This artificial intelligence may be used to keep people safe by spotting handguns in places they shouldn’t be.

Humans are hardwired to quickly spot objects in the corner of our eyes: Kids getting into trouble, snakes slithering into the room, and the like. That task isn’t as easy for algorithms, but they’re getting better. Someday soon they may even help humans spot dangerous objects we might have missed. Researchers at the University of Granada, for example, have used deep learning to train an algorithm to identify handguns from video in real time.

The system may be used in home security systems, police body cameras, and video surveillance at places like schools, airports, and malls. It could also be applied to videos online as a way to monitor and map violent content.

Although around 30 percent of American adults say they own guns; the weapons have a steady presence in our lives. They seem to be everywhere you look — carried by cops on TV or as a key component in many of today’s most popular video games. So it’s appropriate that the University of Granada researchers trained their system on movies like Pulp Fiction, and the Mission Impossible and James Bond franchises, through which many of us also had our first introduction to firearms.

After training the algorithm on low-quality video of these classic films, it was able to effectively identify drawn handguns more than 96.5 percent, while analyzing in real time at 5 frames per second. When the algorithm spots a handgun, it borders it with a red box. The researchers think that the model can be combined with an alarm system in order to quickly and inexpensively identify firearms in places where they’re forbidden or may pose a threat.

Although the scientists don’t expect their model to completely replace current weapon-detection systems like metal detectors, they see it as a complementary system to help address some of the drawbacks of conventional methods. For example, metal detectors require guns to pass through a certain point, which — as anyone whose passed a security checkpoint knows — can create a bottleneck of people in crowded areas. The model is also able to discriminate between guns and other metal objects, which the metal detector cannot.

You’ll need to sell a kidney for Intel’s new 28-core Xeon CPU

Why it matters to you

PC hardware is rarely on the cheap side, but be thankful you don’t have to shell out for the uber-expensive new Platinum Skylake-SP CPUs.

Intel has announced the specifications and pricing information for its new range of Xeon Skylake-SP Platinum series processors, and though the performance should be high, the cost is even more so. The 8180M comes packing 28 cores and 56 threads, with a turbo clock that hits 3.8GHz and a power draw of 205W, but it will set back prospective buyers as much as $13,000.

Xeon processors have always been expensive, but they’re not aimed at consumers or even deep-pocketed hardware enthusiasts. They’re designed with enterprises in mind and because of that and some of their additional features, the price tags tend to be exorbitantly more than their desktop counterparts.

If you compare the 8180M with Intel’s top-end and newly released Core i9-7900X, for example, the specifications differences are huge. While the latter chip sports 10 cores and 20 threads, the former has almost three times as many of each. While that doesn’t necessarily justify the $12,000 different in pricing, it does go some way to explaining it.

Not all of Intel’s 16 new Platinum processors detailed under the new metallic branding, are as expensive of course. The range starts with a CPU with 16 cores at 2.8GHz turbo for $3,155, all the way up to the $13,000 monster. You can save $3,000 at the top end if you ditch the extended DRAM support of the M version of the 8180, which many may choose to do, but few of these CPUs could be considered cheap.

Indeed the top end offering, with its near 40MB of L3 cache, is almost twice as expensive as the top-end Xeon processor from the previous generation (thanks Anandtech). It’s likely that the companies that buy these chips in bulk won’t pay a list price, but even factoring that in, this generation has had a big price jump over its predecessor.

For companies not looking for the out-and-out top-tier performance and flexibility of Platinum CPUs, the Gold range — which may show up in future iMac Pros — isn’t too far off in features. It offers up to 22 cores with a frequency of 3.7GHz on its 6152 CPU. It’s also far, far cheaper, with a price tag of ‘just’ $3,655.

The actual list of Gold options is extensive, if a little confusing as a result. It’s split into the 6XXX series and 5XXX series, with a total of 32 chips between them. That gives manufacturers and businesses a lot of options, though they’ll need to know specifically what they’re looking for, as Intel has options for long-life products, extended DRAM functions, various power draws, and more within the range.

It will be interesting to see if that complexity aids the competition in AMD’s more streamlined Epyc range of CPUs. Time will tell.

Nokia 6 goes up for sale in the U.S. for $229; Prime variant for $179

Nokia 6 is one of the best phones in the sub-$250 segment.

The Nokia 6 is now up for sale on Amazon for $229. The standard variant of the phone is unlocked for use on AT&T and T-Mobile, and Amazon is also offering a variant with lock screen ads that retails for $179, a $50 discount.

The phone has a 5.5-inch Full HD display, and is powered by an octa-core Snapdragon 430 chipset with four Cortex A53 cores clocked at 1.4GHz. There’s 3GB of RAM, 32GB storage, 16MP camera at the back, 8MP front shooter, NFC, and a 3000mAh battery. The Nokia 6 is running Android 7.1 Nougat on the software front, with the interface devoid of any customizations. There’s also a fingerprint sensor up front, and the phone has a hybrid SIM card tray that sees the secondary SIM slot doubling up as a microSD slot.

The Nokia 6 is available in copper and black color options, with the blue and silver variants still limited to pre-orders. If you’re in the market for a sub-$250 phone with decent specs and a near-stock Android experience, the Nokia 6 is one of the best options available.

See at Amazon

Moto E4 Plus with 5000mAh battery launches in India for ₹9,999

Get two-day battery life and a near-stock Android experience with the Moto E4 Plus.

Motorola is expanding its catalog in India, and the latest addition is the Moto E4 series. The Moto E4 Plus will go up for sale later today on Flipkart for ₹9,999, and the smaller Moto E4 will be available at thousands of retail stores across India for ₹8,999.

The highlight of the Moto E4 Plus is the 5000mAh battery, which should easily last two days with moderate use. The phone features a 5.5-inch 720p display, 1.3GHz quad-core MediaTek MT6737 SoC, 3GB of RAM, 32GB storage, two SIM card slots along with a dedicated microSD slot, 13MP rear camera, 5MP front shooter, 802.11 a/b/g/n, and LTE connectivity.

Meanwhile, the Moto E4 sports a 5.0-inch 720p panel, and is powered by the same 1.3GHz MT6797 chipset as the E4 Plus. You get 2GB of RAm, 16GB storage, microSD slot, an 8MP camera at the back, 5MP shooter up front, LTE, and a 2800mAh battery. Both phones run Android 7.1.1 Nougat out of the box.

As is the case with all Motorola phones, there are several launch-day offers available. If you’re an Idea subscriber, you can get 84GB of data (1GB a day) for just ₹443. You will also receive a two-month subscription to Hotstar premium for free, and get ₹850 off on the Pulse 2 wired headset. Flipkart is also offering a buyback guarantee, through which you’ll get up to ₹4,000 when you decide to upgrade.

The Moto E4 Plus will go up for sale starting 11:59 p.m. IST on Flipkart. Who’s interested?

See at Flipkart