How Facebook’s New AI is Thwarting Facial Recognition Programs

Sentiment on facial recognition varies somewhat by region. In Australia, for example, more than two-thirds of the population supports facial recognition, particularly when it’s used by law enforcement. In the UK about 70% concur, though little more than half the people in the US agree. Of course, these statistics relate specifically to law enforcement using the tech. Approval plummets for virtually all other applications, especially when there’s no clear benefit to the general public, such as for marketing purposes or whilst selecting job candidates.

The challenge is, the technology has grown so fast that nearly any company can wield its power, giving major rise to privacy concerns. Facebook, however, aims to do something about it.

Computers Use Facial Landmarks to Identify a Person

Unlike humans, who can more or less identify a face at a glance, software programs have to use mathematical calculations to establish uniqueness. Each person has 80 landmarks, also called nodal points, and when the distance between them is measured, a clear faceprint, or code, emerges. For example, the distance between the eyes represents one measurement, while the shape of cheekbones and length of jawline represent two more.

Facial Recognition Software Can Be Exploited

In many cases, the technology is beneficial. For example, Facebook leverages it to identify faces in photos that are posted, allowing the poster to tag who’s in the photo easier. The company rolled out privacy measures that warn someone when their photo is uploaded and also lets people opt out.

However, systems can be fooled. In the early days of facial recognition, police in Tampa, Florida installed cameras and leveraged the tech in an effort to reduce crime. People were seen wearing masks; foiling the pricey system in the most low-tech way. Researchers at Carnegie Mellon University took it a step further and created specialized glasses that exploit the very way computers read faces—by making the landmarks appear as if they are in different places. These “perturbed eyeglasses” make it possible to impersonate anyone and work roughly 92-100% of the time.

Since then, many ways to impersonate others via facial recognition software have emerged, and the unfortunate consequence is that these techniques can be leveraged for nefarious purposes. For example, someone can be placed at the scene of a crime or “deepfake” videos may be created that depict an individual saying or doing virtually anything.

Facebook is Working on De-Identification Tech

Carnegie Mellon researchers also worked with a technique called dodging, which essentially distorts a person’s landmarks and prevents recognition, but does so in a way that’s not detectable by the human eye. In these cases, a person may be indistinguishable or mistaken for any other person, and by their tests, is effective 100% of the time.

It’s this method that Facebook is working on, but their team has identified how to make it work in video, including live video streams. ““Recent world events concerning the advances in, and abuse of face recognition technology invoke the need to understand methods that successfully deal with de-identification,” the researchers say. “Our contribution is the only one suitable for video, including live video, and presents quality that far surpasses the literature methods.”

Although it’s unclear how Facebook plans to use this tech, they’ve approached dodging or de-identification in a novel way that helps maintain the quality of the images while making it impossible for computers to identify the subject. Going forward, it could lead to further developments that can aid in privacy measures or prevent the creation of deepfakes.

Twitter IPO and the tale of three Peters and a Dick!

Twitter expectedly filed for an initial public offering but what made bigger news was the tale of “three Peters and a Dick,” as the microblogging site got broadsided for its lack of gender diversity and the broader Silicon Valley culture got upbraided for its historical insensitivity.

First the naughty phrase.

The credit for that goes to Kara Swisher of All Things Digital, courtesy an unnamed board member. The reference was to Peter Fenton, Peter Currie, Peter Chernin of Twitter’s governing body and CEO Dick Costolo, who hardly covered himself in glory with his responses.

The row was kicked off by The New York Times, which unearthed from the IPO filing the fact that Twitter didn’t have a single woman on its board. Also, it had just one woman – a recent hire for the position of general counsel – in its senior executive ranks, despite a nearly equal number of users of the social media site.

“The fact that they went to the IPO without a single woman on the board, how dare they?” the rising tech sage Vivek Wadhwa charged in the Times article, instantly emerging as a champion of Silicon Valley women. He also called into question the “elite arrogance of the … Twitter Mafia.”

Costolo responded by calling Wadhwa a “Carrot Top” of academic sources, alluding to the standup comedian. He also defended himself with the following: “I was making fun of his propensity for silly hyperbole. I didn’t say anything about the topic or even reference it!”

Wadhwa, who is crowdsourcing a book on women in technology, responded with a longer article in TechCrunch, backing his views and taking hefty swipes at Silicon Valley’s culture. Consider the passage below:

“Here is the harsh reality: Silicon Valley is a boys’ club — a fraternity of the worst kind. It stacks the deck against women. It leaves out blacks and Hispanics. And it provides unfair advantage to an elite few who happen to be connected. Yes, it is also one of the most diverse places on this planet, where anyone can strike it big. But that opportunity only comes of learning the Valley’s rules of engagement and mastering them. Very few can.”

Down Under, I think we recognise the need for greater gender diversity but I am not sure we hold similarly harsh views of our culture. You tell me if we go far enough to ensure gender balance out here.

The testosterone-soaked row has more or less footnoted the other key elements of Twitter’s IPO. To put it briefly, many are simply underwhelmed by Twitter’s numbers and growth prospects. Given the less than enthusing experience with the Facebook IPO, the outlook is downbeat. More interested readers, and investors, can look up a good look at Twitter’s numbers, compared with those of Google and Facebook at a similar stage ahead of their IPOs, on Businessweek here.

What’s a cat got to do with Facebook news feeds?

Ever wondered why you sometimes get the most inane stuff in your Facebook news feeds but not the most “important” ones?

Put it down to poor machine intelligence. Till last year, machines barely knew a cat from a dog, or even a human.

Thus far, Facebook has used conventional machine intelligence to pare down the 1,500 average updates Facebook users get to a meaningful range of between 30 and 60.

“We have five minutes to really delight you,” Facebook CTO Mike Schroepfer told MIT Technology Review recently, referring to user habits of frequently checking news updates, growingly on mobiles when waiting in queues or even when inside movie halls.

“It’s not that I look at my news feed once at the end of the day; I constantly pull out my phone while I’m waiting for my friend or I’m at the coffee shop. We have five minutes to really delight you,” Schroepfer said.

Facebook is turning to a technology called “deep learning” to raise the quality of its news feeds, raise engagement levels of the billion-plus users of the social network so that it eventually serve better targeted ads. It is following in the footsteps of tech giants such as Google and Microsoft, both of whom have made rapid strides in artificial intelligence.

Google, for example, who else?) set up “a neural network of 16,000 computer processors with one billion connections and let it browse YouTube,” as Wired magazine recounted, to test, primarily, if the machine can recognise a cat or humans by watching videos.

Last year, its system achieved 81.7% accuracy in detecting human faces, 76.7% accuracy when identifying human body parts and 74.8% accuracy when identifying cats. That is the order of difficulty in making computers not just intelligent but also self-learning. Still, Google used its experiments to reduce errors in its voice recognition services.

Microsoft has used deep learning to build a system to translate in real-time Mandarin Chinese to English, and for its Bing search services.

The bottom line: Deep learning represents a quantum leap in machine learning, and offer performance that is three times faster and better than conventional machines. Deep learning systems require less human intervention and their superior performance emanates from their own ability to learn – akin perhaps to the way a human learns.

Facebook plans to use deep learning to spruce up a lot of its services. Besides improving the quality of its news feeds and status updates, Facebook can help people organize their photos or choose the best ones to share with friends, according to Schroepfer.

Srinivas Narayanan, an engineering manager at Facebook who is helping to assemble its new Artificial Intelligence group, told Technology Review the social network can help advance deep learning is by drawing on its recent work creating new types of hardware and software to handle large data sets.

Facebook’s deep learning research is being led by Marc’Aurelio Ranzato, who was hired from Google, and includes Yaniv Taigman, cofounder of the facial recognition startup Face.com, and computer vision expert Lubomir Bourdev.

With a third tech major getting into AI, it might be reasonable to expect some rapid advances for a generation that simply can’t wait!

Does the world need Bitcoins? Or, maybe, something similar

If you haven’t heard about Bitcoins, you live in a planet other than the Internet. In case you still missed it, there’s time to learn of it even though you may not want to buy them – certainly not at their present values.

The so-called currency – created four years ago on the Internet – has been on a tear since the banking crisis in Cyprus broke last month. It sure has had its ups and downs in the past but nothing like the tumults it is witnessing just now – hitting a lifetime high value of $266 and then plunging $160 one fine morning (April 10).

What are Bitcoins?

Somewhere on the Internet, a man – a likely hacker called Satoshi Nakamoto, but never really identified – dreamt up the digital currency as an alternative to currencies controlled by government. He designed to be like the Internet – free, global and unregulated. But what is good for a network like the Internet may not necessarily be good for a currency.

Even in its short history, the Bitcoin has less disrupted the world’s monetary system – the value of all Bitcoins in circulation has been a mere $3 billion, at its peak valuation – and gotten more disrupted by hackers and speculators. Yes, it’s being accepted on many Internet sites – and even a Manhattan bar – but it’s presumptive to call it a currency.

Economists like Paul Krugman, kindly, consider the system no better than the gold standard, when it’s probably a lot worse. That’s because Nakamoto has capped the number of Bitcoins that can be “mined” at 21 million – a figure likely to be reached by 2032. Many think Bitcoin was doomed at birth because of this irrational limit on the currency, not to mention the manner in which it is created – Bitcoins can be “mined” at a rate of 2.5 coins per minute, given current computing ability. That is a reason why some think quantum computing could enable Bitcoins to be produced at a much faster rate. But then the absolute cap still exists.

Regardless, there is much enthusiasm among hackers and others to create a global currency that is less regulated, and free from recurrent financial crises. And, guess what, entrepreneurs and venture capital seem to share that enthusiasm, notably after the recent surge in the value of the Bitcoin.

The Winklevoss twins – Tyler and Cameron – are believed to have hoarded Bitcoins and could even be among the largest holders of the virtual currency. For those not in the know, the twins are co-founders with Mark Zuckerberg of the earliest version of Facebook, and extracted a fat cheque from Zuckerberg.

While that only feeds the raging speculation, there is a bid to create something as good as Bitcoin. Behind that is a company called OpenCoin, and behind that company is Netscape founder Marc Andreessen, now a successful venture capitalist.

OpenCoin’s goal is simple: build a better Bitcoin. Who could object to it?

Facebook Home, or social networking 24/7

Do we need more Facebook? Or less?

To many of us it may be a no-brainer, given the fair bit of fatigue we encounter with social networking.

But that is not how Mark Zuckerberg thinks. If anything, he believes Facebook has many more worlds to conquer, so to say. He reckons only a third of the world’s population is on the Internet – the rest of the population still numbers over four billion. What’s more, the rest is going to debut on the Net via smartphones, not PCs.

Hence Facebook Home. Check out a hands-on video of the app here at TechCrunch.

For over two years, speculation had been rife about a Facebook Phone. But Zuckerberg has surprised most of us by his choice of tools for social networking domination – a humble app, rather than a full-fledged mobile phone. But Zuckerberg is not a smart billionaire several times over for nothing. Look closely, and you discover that the Facebook Home app, like Zuckerberg’s ambitions, is far from humble.

Facebook Home is an in-your-face app that virtually takes over your cellphone, converting it into a smart Facebook hub. It makes it easier, and faster, to check Facebook – shaving seconds off an activity that apparently consumes 20% of our mobile time, a stat from Facebook.

Zuckerberg hasn’t quite explained why he shied away from the initial Facebook Phone but he clearly articulated the way he created Facebook Home.

“We asked ourselves — if we’re already spending this much time on our phones, how can we make it easier? What if they were designed around people first, and you could also just happen to interact with apps?” said Zuckerberg, who dropped out of Harvard University to start Facebook.

Here are some key features of Facebook Home:

• Facebook Home’s home screen delivers a constant stream of full-screen photos and other information – even when the phone is locked
• Home also has a new more efficient notification system.
• A Chat Heads feature, a pop up that disrupts whatever else you are doing, gives Facebook a primacy many might want but, arguably, many might simply hate.

So, is Facebook going to own the world?

Unlikely, but it will take a bigger slice of our time, and the social networking pie, for sure. Facebook Home is available only on Google’s open source Android operating system. Technically, Facebook Home tweaks native features of Android to its need, just as Amazon has done on its range of Fire tablets. But Apple, Blackberry or Windows wouldn’t allow any of its apps to carry out such major tweaks of its integral features. Conceivably, then, Facebook Home’s reach could be limited to some mobile operating systems, and certainly from Apple’s highly popular iOS.

Arguably then, Facebook Home’s future is tied firmly to that of Android. But that is, right now, a winning horse, steadily gaining global market share, thanks in the main to Samsung. Also, Facebook may be able to recreate Home on some other open source mobile operating systems (Mozilla’ Firefox or Ubuntu Touch), providing it with a gateway to the Third World, the countries primed for a smartphone explosion.

Just over a year since turning its attention to the mobile, Facebook may have a winner on its hands!

Twitter’s Wall Street run may not quite make history

A week after Twitter celebrated its seventh birthday, reports are abuzz of its likely IPO next year. One analyst firm, Greencrest Capital, cites “management changes, site enhancements, global expansion and new strategic alliances” as proof of the company’s Wall Street run.

If and when it happens, Twitter’s initial public offering should be a signal event for the tech company, only the third after Google and Facebook in this young millennium to change our world. In short bursts of 140 characters, Twitter lights up our lives everyday with news, scandals and much more. When it’s not saving lives in Haiti or Mumbai, it reshapes history.

It foments rebellion in Iran and regime change in Egypt, and just as lightly brings in celebrity gossip in the afterhours. It seems nothing could be beyond Twitter’s sweep.

Yes, Twitter is all about changing the world but Twitter is unlikely to cause any such thing on Wall Street, certainly not after Facebook’s act last year. To start with, Twitter will likely be valued at about $11 billion. That’s tiny when compared with Facebook’s opening day valuation of $104 billion. For a company that took a long time to come up with a compelling business model, that may not be bad. Also, it might have significant upside – revenue upside in mobile, and social impact in television.

Like Google and Facebook, Twitter is just discovering mobile ad growth. Its mobile ad revenues are projected by Greencrest to touch $382.8 million in 2014 – a nearly three-fold jump from $134.9 million in 2012. Total revenue in 2014 is projected to be just over $1 billion. eMarketer, another analyst firm, estimates 2014 ad revenue to be $950 million, up from the estimated $604 million this year.

To Twitter, television offers scope for greater impact. We have seen growing synergies between social media and TV. A U.K. survey, for example, identified television as the most popular topic to tweet and found that 40% of evening tweets were about television. Given the emerging scenario, Twitter has been smart in signing up with TV ratings leader Nielsen Media for creating a Nielsen Twitter TV Rating System.

Ultimately, Twitter will find thriving revenue streams and be a successful listed company. But the microblogging services will always be more about changing the world, much less about laughing all the way to the bank.

Does technology need a feminist?

Sheryl Sandberg’s book, “Lean In: Women, Work, and the Will to Lead,” is Amazon’s Book of the Month, and the buzz phrase used by Facebook’s chief operating officer is the latest mantra for career-minded women.

Sandberg’s broad analysis is this: Women give up on their careers too soon, primarily to beat the biological clock and have kids before it’s too late. Instead, she suggests, they should work really hard and set the stage for a return after giving birth to kids.

Preditably perhaps, she has been greeted with howls of protests, notably by (who else?) other women.

Take Allison Pearlson in The Daily Telegraph: “I suspect history will relate that parents who neglected their young in pursuit of their ambitions caused more problems for society in the 21st century than those slackers who failed to “lean in” to their careers.

“Sheryl Sandberg’s colleagues ran a sweepstake on how long it would take their boss to be back on email after the birth of her first child. She lasted a whole day. Poor baby.”

To be fair, Sandberg has walked the talk. She worked at McKinsey, worked for the poor and ill in India, served as the chief of staff to Treasury Secretary Lawrence Summers, took in an MBA from Harvard Business School, and laboured through two children while holding down a high-profile position at Google before landing at Facebook.

Still, two questions remain.

One, women in technology might do with a role model but do they need a feminist?

Sandberg is on a lot of power lists but not the only one in technology. There is Marissa Meyer at Yahoo and Meg Whitman at Hewlett Packard. Technology attracts some of the brightest women, and in today’s day and age, it’s vain to think they need help figuring out work-life balance, or feminist advocacy.

Two, does she inspire career women across the professions and the world?

Early criticisms are ominous. Writing in Forbes, Meghan Casserly, said: “Taking advice from a woman who can make her own rules doesn’t compute as a young woman who operates firmly under the thumb of her employer. Sure, it’s something to aspire to—but in the same way I aspire to buy $99 t-shirts in bulk like Gwyneth Paltrow so I never have to worry about laundry-day.”

And then there is the harsher comment from the Telegraph’s Parlson: “The difficulty with a successful woman setting out to write a book about work and ambition is that half her target audience won’t know what she’s talking about because they’re too busy trying to make ends meet. The other half will hate her because she’s successful.”

It would be a pity if Sandberg’s message is already lost on women.

f

Future is social. Just see TV

Every day, Australians spend about three hours in front of their televisions. That number hasn’t changed significantly over the past five years. But what has changed – imperceptibly, perhaps – is the way we watch TV.

I know we are still waiting for Apple’s iTV or something similar to dramatically transform our living room experience. But society doesn’t always sit and wait for technology innovations. It imaginatively uses what is available. In the case of television, what we seem to have done is married it with social media and created what seems like a perfect union.

Increasingly, Australians are watching TV not in isolation, but with social media as a companion – sharing our experiences with our loved ones and friends. So much so, nearly one in two Australians is snuggling in front of the TV with a smartphone or tablet in hand, and actively tweeting or posting status updates on Facebook. What’s more, they are using social media to discover the best in TV.

This knowledge of our changing behaviour comes from two separate sources. The first to come was Yahoo!7’s second annual survey. Its two major findings, based on a survey of 7,741 Australians, were:

• 43% watch TV with a remote in one hand, and social media in the other; and
• 42% pick the shows to watch based on social media recommendations, a nearly three-fold growth from 16% in 2011.

The second source is Online Circle, a digital media agency in Melbourne that manages Facebook impact for corporates. It found that television programmes soared to the top of the charts on the strength of social media fans. For the first time since Online Circle began measuring industry-wide data for Facebook impact, it found a growing, but not perfect, correlation between the number of fans and ratings of television programs.

Overall, programs on Australian television had nearly 9.4 million Facebook fans. The children’s show, “Bananas in Pyjamas,” was the single most popular one with two million fans. “Home and Away” and “Masterchef Australia” were also in the top five.

Television is not the only medium that is making waves in social media. It’s radio too, where Online Circle reports an incredibly high engagement rate – 15.6%, compared to television’s relatively modest 6.9%. That is explained by the fact that radio, far more than television, revolves around community and is best experienced when shared, which is where the social media element adds the X-factor.

Going forwards, expect TV and radio to engage even more on social media. But the message is broader for companies still grappling with Facebook and the likes. The future is social!

Social media Big Brother coming

Three years ago, then Google CEO Eric Schmidt announced that privacy is dead in the Internet era. So did Facebook founder Mark Zuckerberg. Privacy watchdogs fought the heads of the two Internet giants but unknown to all of us a more real, more scary Big Brother seems to have emerged.

At about the same time as Schmidt and Zuckerburg were mouthing off, American defence contractor Raytheon quietly – unknown to most – made a presentation to government agencies of a software program that could track down anybody using digital footprints from social media and other sources. The Guardian daily posted a video of the presentation that should scare the daylights out of private citizens. It reveals the story of the software, demonstrating how Raytheon tracked one of its own employees to establish its strength.

Raytheon’s social networking-tracking program – called Rapid Information Overlay Technology, or RIOT – uses Facebook, Twitter and the location-based Foursquare to dig up a person’s every move. It apparently uses a hidden tag in mobile posts called “exif” files to zero-in on latitudes and longitudes, helping pinpoint the physical location of the user. Some reports suggest governments could, using this software, assess person’s “behaviour history” to predict future actions.

Obviously, the whole thing is scary for most. It is one thing to have private companies to track our every move, primarily for profit, and quite another when governments begin to use software such as RIOT to spy on us. When, once, a website revealed Schmidt’s home address, the Google chief – whose privacy policy is ambivalent, not to mention dubious – reacted angrily. Nobody likes invasions of privacy.

When reports about RIOT began to appear last month, Raytheon was quick to say the software was only for demonstration, and that it has never been sold to any agency. But then the demonstration was three years ago. Raytheon likely has continued development of the software, and at some point will sell it to U.S. agencies. For American citizens, it’s sure a scary thought. For the rest of us, it should still be scary because some other company is going to come up with something similar and sell it to our governments. Imagine what such software can do in the hands of the rules of Egypt, Zimbabwe, North Korea or Iran!