Scan Anything and Let Your Phone Do the Rest

New discovery: With a new app from Digimarc, someone could scan photos from a newspaper or other media to pull up more information—provided those images are in the company’s database.


Designed to teach math to students in poor countries, the device will be the first to use a new energy-efficient computing strategy.

Many people rely on their smart phones to search for things online. At the movies, users might try to identify an actor from a film trailer. At a concert, they might hear a song and check which album it was on. When shopping, they might try to find the best deal on a product by searching nearby stores. Apps that identify songs, images, and video, or that read barcodes, make it easier to do this.

Now Digimarc, based in Beaverton, Oregon, has combined these functions into Discover, a single app designed to identify input from a person's environment and pull up related information.

Similar to apps like Shazam, SoundHound, or Barcode Scanner, Discover uses a smart phone's camera and microphone to "capture" a sample of audio or an image, then identifies it through Digimarc's own database and searches for related material online.

Unlike these apps, though, Discover combines a variety of media search functions into a single app that will allow users to scan images, audio, video, and even barcodes or QR codes (two-dimensional versions of barcodes)—all without switching between apps. Discover is available for free on both iOS and Android phones.

However, Discover's usefulness is limited by the number of companies that utilize the system. As of yet, this system is only implemented by a small number of publications.

Discover requires "digital watermarks" to identify images and video. These work much like the watermarking used on currency or official documents, by inserting a transparent image on top of another image. The difference is that digital watermarking is specifically designed to be recorded and decoded by software.

Digimarc offers an online service that advertisers and companies can use to purchase and place watermarks in their ads and images. Then, when a user scans one of those images with a watermark, the Discover app searches through Digimarc's database. Once the app has identified an item through the database, it searches online and brings up related 

While Discover's image and video search functions only work with the company's own digital watermarking system, its audio search takes advantage of the information stored by the digital media company Gracenote. Barcode searches, however, use general Web searches to locate relevant data.

Tony Rodriguez, chief technology officer at Digimarc, believes that the innovative aspect of the app is that it not only consolidates several search technologies, it uses them concurrently. He believes that the future of mobile search apps lies in this direction. Rodriguez says that Digimarc hopes to work with partners who will expand on Digimarc's vision. He says, "Our goal isn't to be the end-all platform, but to add to a growing group."

But not everyone feels that Discover is a necessary tool for consumers. Sean Owen, one of the creators of Barcode Scanner, the most popular barcode and QR code scanner for Android, feels that even his own app "is a bit of a novelty and niche on smart phones—and video and audio search even more so." While Owen thinks the idea of a consolidated app is "very cool," he feels that consumers, including himself, would rather "choose individual best-of-breed apps" rather than one app that doesn't perform as well, but covers multiple tasks.

Stephen O'Grady, principal analyst at RedMonk, feels that Digimarc's decision to use its own database and digital watermarking system will hold it back from reaching the mainstream. And while consolidated search apps may become popular in the long term, he warns, "Over the short term, we're talking about changing the core navigating behavior of hundreds of millions or billions of users."

With Big Data Comes Big Responsibilities

Double trouble: Researchers says big data sets can reveal unintended, questionable insights, especially when combined, such as with the recent blend of information from Spotify and Facebook.

The reams of data that many modern businesses collect—dubbed "big data"—can provide powerful insights. It is the key to Netflix's recommendation engines, Facebook's social ads, and even Amazon's methods for speeding up the new Web browser, Silk, which comes with its new Fire tablet.

But big data is like any powerful tool. Using it carelessly can have dangerous results.

A new paper presented at a recent Symposium on the Dynamics of the Internet and Society spells out the reasons that businesses and academics should proceed with caution. While privacy invasions—both deliberate and accidental—are obvious issues, the paper also warns that data can easily be incomplete and distorted.

"With big data comes big responsibilities," says Kate Crawford, an associate professor at the University of New South Wales, who was involved with the work. "There's been the emergence of a philosophy that big data is all you need," she adds. "We would suggest that, actually, numbers don't speak for themselves."

Crawford's paper, written with Microsoft senior researcher Danah Boyd, illustrates the ways that big data sets can fall down, particularly when used to make claims about people's behavior. "Big data sets are never complete," Crawford says. For example, researchers often study Facebook to analyze people's social relationships, using connections made through the social network as a stand-in for real-world ties. But it's common for Facebook to show a distorted picture of people's closest social relationships, such as with parents, live-in romantic partners, or friends seen daily. "Facebook is not the world," Crawford says.

Google is a poster child for the power of data. The company has transformed a massive amount of information, gathered through its search engine, into a commanding ad network and powerful role as the gatekeeper of much of the world's information.

At a conference on Knowledge Discovery and Data Mining in August, I watched Google's director of research, Peter Norvig, demonstrate the true power of a large data set, using the example of machine translation. Norvig showed that training algorithms on very large data sets, like those it has collected from the many Web pages it crawls that are available in multiple languages, can produce dramatic results. With enough data, Norvig said, even the worst algorithm performs far better than what can be achieved with a smaller data set.

But Crawford and Boyd's work shows that studying large data still requires finesse. Twitter, which is commonly scrutinized for insights about people's moods, attitudes toward politics, and other aspects of daily life, presents a number of problems, the researchers say. About 40 percent of Twitter's active users sign in to listen, not to post, which, Crawford and Boyd say, suggests that posts could come from a certain type of person, rather than a random sample. They also note that few researchers have access to all Twitter posts—most use smaller samples provided by the company. Without better information about how those samples were collected, studies could arrive at skewed results, they argue.

Crawford notes that many big data sets—particularly social data—come from companies that have no obligation to support scientific inquiry. Getting access to the data might mean paying for it, or keeping the company happy by not performing certain types of studies.

The researchers add that big data can also raise serious ethical concerns.

Many times, Crawford notes, combining data from different sources can lead to unexpected results for the people involved. For example, other researchers have previously shown that they can identify individuals by using social media data in combination with supposedly anonymized behavioral data provided by companies.

Jennifer Chayes, managing director of Microsoft Research New England, says her lab has had firsthand experience with such problems. The lab wanted to run a contest for researchers to analyze a set of search data, she says, and was going over the data carefully to avoid the sorts of deanonymizing scandals that have occurred from search data releases in the past. They discovered that people often entered search terms that were personally identifying and embarrassing—such as, "Is my wife Jane Doe cheating on me?" The lab nixed the contest. Chayes says, "We began to realize how much we didn't understand about human behavior around search engines."

Handling big data sets takes almost impossible care, agrees Alessandro Acquisti, an associate professor at Carnegie Mellon who has studied the unintended information that data sets can reveal. Even public data sets raise questions, such as what to do with information that people post and then subsequently want to delete, he says.

Given the quantity of information now available on the Internet, Crawford argues, researchers need to slow down and think about the methods they use. "[The effect of the availability of big data] did shock a lot of people," she says. "And it should."

How Intel Turbo Boost Works

Normally the computer processor in your laptop or desktop has a standard clock speed which partially determines how quickly it performs. While the processor might lower its clock speed at times in order to conserve power, the clock speed which is stated when you buy the computer is the fastest clock speed you’ll receive unless you decide to overclock.

If you do decide to overclock, or you ever speak to someone who regularly overclocks processors, you’ll discover a dirty little secret ““ the clock speed a processor ships at is typically much lower than the actual maximum clock speed which the processor could achieve.

The extra headroom isn’t used only because the manufacturer (Intel or AMD) needs to plan for worst case scenarios, which means they need a processor which is sold as a 3GHz processor to work at that speed even if someone decides to use a winter jacket as a PC case.

At least, that is how processors used to be. However, Intel’s new Core i5 and Core i7 processors have a feature called Turbo Boost which has the ability to dynamically scale up the clock speed of a processor depending on the thermal headroom available.

How Intel Turbo Boost Works

Intel Turbo Boost monitors the current usage of a Core i5 or i7 processor to determine how close the processor is to the maximum thermal design power, or TDP. The TDP is the maximum amount of power the processor is supposed to use. If the Core i5 or i7 processor sees that it is operating well within limits, Turbo Boost kicks in.

Turbo Boost is a dynamic feature. There is no set-in-stone speed which the Core i5 or i7 processor will reach when in Turbo Boost. Turbo Boost operates in 133Mhz increments and will scale up until it either reaches the maximum Turbo Boost allowed (which is determined by the model of processor) or the processor comes close to its maximum TDP. For example, the Core i5 750 has a base clock speed of 2.66GHz but has a maximum Turbo Boost speed of 3.2GHz.

However, Intel still advertises these processors by their base clock speed. This is because Intel does not guarantee that a processor will ever hit its maximum Turbo Boost speed. I have yet to hear of an Intel processor which can’t hit its maximum Turbo Boost speed, but hitting the maximum Turbo Boost is dependent on workload ““ it won’t happen all of the time.

Why Turbo Boost Rocks

Despite Turbo Boost’s lack of predictability, it is still an excellent feature. It provides a solution to the problem of compromising between dual and quad core processors.

Before Turbo Boost the choice of purchasing a dual core or quad core processor was a compromise. Dual core processors were clocked faster than quad core processors simply because having more cores increases power consumption and heat generation. Some programs, like games, favored dual core processors, while other programs, like 3D rendering software, favored quad cores. If you used both types of applications you had to make a choice about which was most important to you. You couldn’t receive maximum performance in both from a single processor.

Turbo Boost gets rid of this compromise. If you use the Core i5 750 in a 3D rendering application it will probably only operate at its base clock speed because all four cores will be used. However, if you use the Core i5 750 with a game which only needs two cores ““ presto! – the third and fourth cores go into a low power state and the two cores you’re actually using are running at a clock speed as fast as what you’d expect from a standard dual core processor.

The Future of Intel Turbo Boost ““ And AMD’s Response

Turbo Boost is a great feature, and it is part of the reason why Intel’s latest processors are often superior to those from AMD. However, there is still more potential to be tapped. By the end of 2010 Intel will have released ultra-low voltage Core i5 and i7 processors for laptops. These processors will use Turbo Boost as a way of improving battery life.

For example, Intel will be releasing a processor called the Core i7 620UM. This processor has a base clock speed of only 1.06GHz. However, it has a maximum Turbo Boost of 2.133 GHz. What we will end up with is a processor which will run at only the base clock when on battery but can double its speed when plugged in.

Intel’s success with Turbo Boost has not gone unnoticed by AMD, however. With the release of the six-core AMD processors, such as the Phenom II X6 1090T, AMD has introduced a similar feature called Turbo Core. Turbo Core isn’t as advanced as Intel’s Turbo Boost, but it is a clear sign of the direction processors will be taking in the future.

It appears the days of set-in-stone processor clock speeds are over. The future will be about changing a processor’s performance on the fly to meet the demands of the user.

Did this article help you understand more about Turbo Boost and why you need it? Still not sure about something? Go ahead and get it answered in the comments.

Searching for New Ideas

Google's head of research explains why artificial intelligence is crucial to the search company's future.

If anyone can preview the future of computing, it should be Alfred Spector, Google's director of research. Spector's team focuses on the most challenging areas of computer science research with the intention of shaping Google's future technology. During a break from aNational Academy of Engineering meeting on emerging technologies hosted by his company, Spector told Technology Review's computing editor Tom Simonite about these efforts, and explained how Google funnels its users' knowledge into artificial intelligence.

TR: Google often releases products based on novel ideas and technologies. How is the research conducted by your team different from the work carried out by other groups?

Spector: We also work on things that benefit Google and its users, but we have a longer time horizon and we try to advance the state of the art. That means areas like natural language processing [understanding human language], machine learning, speech recognition, translation, and image recognition. These are mostly problems that have traditionally been called artificial intelligence.

We have the significant advantage of being able to work in vitro on the large systems that Google operates, so we have large amounts of data and large numbers of users.

Can you give an example of some AI that has come out of this research effort?

Our translation tools can now use parsing—understanding the grammatical parts of a sentence. We used to train our translation just statistically, by comparing texts in different languages. Parsing now goes along with that, so we can assign parts of speech to sentences. Take the sentence "The dog crossed the road": "the dog" is the subject, "crossed" is a verb, "the road" is the object. This makes our translations better, and it's particularly useful in Japanese.

Another example is Fusion Tables, which is now part of Google Docs [the company's online office suite]. You can create a database that is shared with others and visualize and publish that data. A lot of media organizations are using it to display information on Google Maps or Google Earth to explain situations to the public. [During the recent hurricane Irene, New York public radio station WNYC used Fusion Tables to create an interactive guide to evacuation zones in the city.]

Does Google have a particular approach to AI?

In general, we have been using hybrid artificial intelligence, which means that we learn from our user community. When they label something as having a certain meaning or implication, we learn from that. With voice search, for example, if we correctly recognize an utterance, we will see that it lead to something that someone clicked on. The system self-trains based on that, so the more it's used, the better it gets.

Spelling correction for Web search uses the same approach. When Barack Obama ran for president, people might not have been sure how to spell his name and tried different ways. Eventually they came across something that worked, then they clicked on the result. We learned then which of the spellings was the one that got the results, which allowed us to automatically correct them.

We think Fusion Tables will also help our systems learn. If there are thousands of tables that say there are 50 states in the Union, there are probably 50 states in the Union. And the Union probably has states. Don't underestimate that. It sounds trivial, but computers can induce lots of information from many examples.

What new directions is the research group exploring at the moment?

We're looking at projects in security, because it's an increasingly important topic across computing. One area we're looking at is whether you can constrain the programs that you use to work on the most minimal amount of information possible. If they went wrong, they would be limited in what harm they could do.

Imagine you're using a word processor. In principle, it could delete all of your files; it's acting as you. But what if when you started your word processor, you gave it only a single file to edit? The worst it could do would be to corrupt that file; the damage it could do would be very limited. We're looking if we could tightly constrain the damage that could be done by faulty programs. That's an old line of thought. People have thought of this for years. We think it might be practical now.

Google is working hard on its social networking project, Google+. Do you expect your research to contribute to that effort?

Being useful in the social realm is very strong for many of the things that we do. Google+ is a communication mechanism, and we do research on AI problems that could aid communication—for example, how to recommend content, or how to communicate across languages. Ideas like those could help people communicate across their social network.

Google+ also provides us lots more opportunity to learn from our users. Take the "+1" button, for example. That's a very important signal that could be quite relevant to improving how we understand what matters to you. If your 10 friends think something is great, it's very likely you would like to see it.

Movies, Music, and More Move Inside Facebook

Mark Zuckerberg announces ways to listen to music and watch movies inside Facebook, giving users less reason to spend time outside the site. 

New features coming to Facebook this week will let users listen to music, watch movies, and read news without ever leaving the social network's borders. They will also automatically broadcast what users are listening to, watching, and reading, if the user gives permission. The changes, part of an attempt to encourage people to share more of their lives through Facebook, were announced by company founder Mark Zuckerberg at the company's F8 event in San Francisco today.

Zuckerberg announced during his keynote speech that a slew of media companies will collaborate on the project. If a person installs a Facebook app from any of those companies, his or her activity will be shown to friends via a new "ticker" box at the right-hand side of all Facebook pages. Users will be able to click on updates to get access to the same content, which will play, or become viewable, inside Facebook."You're going to discover lots of new things that your friends are already doing all the time right now," said Zuckerberg. "If I see that my friend is listening to something, I can hover over it and just play it; I'm listening with my friend, and my music is synched up with theirs."A similar change will make it possible to watch TV shows or movies that friends are watching through services including Hulu and Netflix. Updates will be sent to Facebook even if a person is using a mobile app or watching through the service's own website. "We think this is going to make it so people can express an order of magnitude more than they can today," said Zuckerberg.Zuckerberg's keynote was his first major public appearance since the launch of Google's competing social network, Google+, and the release of the Oscar-winning movie The Social Network, which painted him in a less-than-flattering light. Saturday Night Live comedian Andy Samberg opened the proceedings, taking the stage in character as his comic version of Zuckerberg. Samberg announced a number of spoof Facebook features, including the "slow poke," before the real Mark Zuckerberg ran on stage for a few minutes of banter. Zuckerberg's presentation was less well rehearsed than a Steve Jobs keynote, but he came across as more accessible than Apple's ex-CEO. The new features may prove controversial. In some ways they resemble Beacon, a failed project from 2007 in which sites like Amazon automatically posted updates to Facebook when a person bought something. Beacon was cancelled after public protests over a lack of privacy controls.Zuckerberg didn't mention Beacon, but privacy issues came up when Netflix CEO Reed Hastings took the stage to demonstrate how people will be able to use his service through Facebook. Hastings explained that the feature couldn't yet launch in the United States because of a decades-old law intended to prevent video rental stores from sharing titles rented by their customers, but he expressed hope that a bill currently before Congress would soon alter that law.Facebook will compile a stream of snippets about a person's listening, viewing, and reading habits into summaries that their friends can read. "Sometimes you discover things your friends are doing right now; other times you want to look at patterns that build up over a period of time," said Zuckerberg. He added that the new features should make Facebook a driving force in making content, particularly music, pay in the digital era.

"The key to making music work [online] is not trying to block you from sharing songs you've bought; it's helping you discover music so you'll buy more," Zuckerberg said before introducing Daniel Ek, CEO of the music service Spotify, who claimed that data from his company's users showed that helping people sample songs makes them "twice as likely" to buy music.

Referring to Netflix, Spotify, and other partner companies planning to integrate with Facebook's new features, Zuckerberg said, "These companies are not just coming up with ways to make movies and TV more social; they're rethinking entire industries." The Web's most-read news source, Yahoo, is also working with Facebook, and News Corp.'s iPad-only magazine, The Daily, will become available through the site.The new features are made possible by technology called the Open Graph, which connects Facebook's data with outside sources of information. First announced in 2010, the Open Graph made it possible for users to recommend books, movies, and Web pages via the now-ubiquitous Like button and have those recommendations appear on their Facebook page. The Open Graph can also describe people's connection to something—for example, showing that they "read," "watched," or "cooked" it. This kind of connection underpins Facebook's new features.When the Open Graph project was first announced, it was pitched as way to allow users' social connections to follow them around the wider Web. The features announced today work in the other direction, bringing activity and content inside Facebook and potentially obviating the need to spend online time outside it.Another new feature, dubbed Timeline, allows users to curate an interactive record of their life over the years. Timeline automatically summarizes a person's past Facebook activity and attempts to identify the most significant moments. Users can also specify particular photos, events, or other content to highlight on the timeline.

I want to search more securely with Secure Sockets Layer (SSL) search - SSL Search

I want to search more securely with Secure Sockets Layer (SSL) search - SSL Search

Quantum Processor Hooks Up with Quantum Memory

Connecting the two could make it possible to perform complex calculations that are far beyond the power of conventional computers.





Super cool: When chilled almost to absolute zero, this chip becomes a quantum computer that includes both a processor (the two black squares) and memory (the snaking lines on either side).

Researchers at the University of California, Santa Barbara, have become the first to combine a quantum processor with memory that can be used to store instructions and data. This achievement in quantum computing replicates a similar milestone in conventional computer design from the 1940s.

Linking a processor and memory elements brings such applications closer, because it should make it more practical to control and program a quantum computer can perform, says Matteo Mariantoni, who led the project, which is part of a wider program at UCSB headed by John Martinis and Andrew Cleland.

Although quantum computing is now mostly a research subject, it holds out the promise of computers far more capable than those we use today. The power of quantum computers comes from their version of the most basic unit of computing, the bit. In a conventional computer, a bit can represent either 1 or 0 at any time. Thanks to the quirks of quantum mechanics, the equivalent in a quantum computer, a qubit, can represent both values at once. When qubits in such a "superposition" state work together, they can operate on exponentially more data than the same number of regular bits. As a result, quantum computers should be able to defeat encryption that is unbreakable in practice today and perform highly complex simulations.

The design the researchers adopted is known as the von Neumann architecture—named after John von Neumann, who pioneered the idea of making computers that combine processor and memory. Before the first von Neumann designs were built in the late 1940s, computers could be reprogrammed only by physically reconfiguring them. "Every single computer we use in our everyday lives is based on the von Neumann architecture, and we have created the quantum mechanical equivalent," says Mariantoni.

Qubits can be made in a variety of ways, such as suspending ions or atoms in magnetic fields. The UCSB group used more conventional electrical circuits, albeit ones that must be cooled almost to absolute zero to make them superconducting and activate their quantum behavior.  They can be fabricated by chip-making techniques used for conventional computers. Mariantoni says that using superconducting circuits allowed the team to place the qubits and memory elements close together on a single chip, which made possible the new von Neumann-inspired design.The only quantum computing system available to buy—priced at $10 million—lacks memory and works like a pre-von Neumann computer.

The processor consists of two qubits linked by a quantum bus that enables them to communicate. Each is also connected to a memory element into which the qubit can save its current value for later use, serving the function of the RAM - for random access memory - of a conventional computer. The links between the qubits and the memory contain devices known as resonators, zigzagging circuits inside which a qubit's value can live on for a short time.

David Schuster leads a group at the University of Chicago that also works on quantum computing, including superconducting circuits. He says that superconducting circuits have recently proved to be comparatively reliable. "One of the next big frontiers for these techniques now is scale," he says. By replicating the Von Neumann architecture the UCSB team have expanded that frontier.Mariantoni's group has used the new system to run an algorithm that is a kind of computational building block, called a Toffoli gate, which can be used to implement any conventional computer program. The team also used its design to perform a mathematical operation that underlies to the algorithm with which a quantum computer might crack complex data encryption.Mariantoni agrees. "We can easily scale the number of these unit cells," he says. "I believe that arrays of resonators will represent the future of quantum computing with integrated circuits."

That's not to say that quantum computers must all adopt that design, though, as conventional computers have. "You could make a computer completely out of qubits and it could do every kind of calculation," says Schuster. However there are advantages to making use of resonators like those that make up the new design's memory, he says. "Resonators are easier and more reliable to make than qubits and easier to control," says Schuster.

Stabilizing Video

A new system cleans up shaky amateur footage.






Steady hand: New software from Google focuses on key elements of a video to stabilize footage. Credit: Matthias Grundmann
Source: "Auto-Directed Video Stabilization with Robust L1 Optimal Camera Paths"

Matthias Grundmann et al.

Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, Colorado Springs, Colorado, June 21-23, 2011

Results: Google researchers have developed a technique that removes the effects of an unsteady filming hand on amateur video. After showing that they were able to smooth out footage in real time while maintaining focus on key elements in the image, they used their algorithm as the basis for a stabilizer application that runs in real time at www.youtube.com/editor. YouTube, the largest site used for sharing amateur video, did not previously have a stabilizing feature in its nine-month-old Web-based video editor.

Why it matters: Shaky footage has always been a mark of amateur filmmaking, since amateurs usually lack the costly equipment that professionals use to stabilize their cameras. Some algorithms are available to clean up this effect, but those typically remove jerking effects without correcting others, such as the slow bounce of a camera held by a person walking.

Methods: The algorithm begins by identifying key objects in the image and using them to plot the path traveled by the camera. It then determines a "best path," which represents the smoothest course for the camera to have traveled. By cropping the frames, it is able to adjust the footage so that the camera appears to have traveled the best path. The algorithm uses tools such as face detection to make sure it doesn't remove key components of the video in the process. Because the computational work is distributed among many machines, the system is fast enough for editing in a browser in real time.

Next Steps: Right now, the user has to specify what size the system should crop the frames to; the algorithms find the optimal path for the given size. In the future, the researchers plan to adjust the system so that it can calculate the ideal size on its own.