The Face Detection Algorithm Set To Revolutionize Image Search: ‘Deep Dense’ Software Spot Faces In Images — Even If Partially Hidden, Or UPSIDE DOWN
The Massachusetts Institute of Technology (MIT) Review, had a February 16, 2015 article with the title above. The subtitle, goes as follows: “The ability to spot faces from any angle, and even when partially occluded, has always been a uniquely human capability. Not anymore.”
“Back in 2001,” MIT notes, “two computer scientists, Paul Viola, and Michael Jones, triggered a revolution in the field of computer face detection. After years of stagnation, their breakthrough was an algorithm that could spot faces in an image in real-time. Indeed, the so-called Viola-Jones algorithm was so fast and simple, that it was soon built into standard point and shoot cameras. Part of the trick, was to ignore the much more difficult problem of face recognition; and, concentrate only on detection. They also focused only on faces viewed from the front, ignoring any seen from an angle. Given these bounds, they realized the bridge of the nose usually formed a vertical line that was brighter than the eye sockets nearby. They also noticed that the eyes were often in shadow; and, so formed a darker, horizontal band.”
“So, Viola and Jones built an algorithm that looks first for vertical, bright bands, in an image that might be noses; it then looks for horizontal dark bands that might be eyes; and then, looks for other general patterns associated with faces,” MIT wrote.
“Detected by themselves, none of these features are strongly suggestive of a face. But, when they are detected one after the other, in a cascade, the result is a good indication of a face in an image. Hence, the name of this process: a detector cascade. And, since these tests are all simple to run, the resulting algorithm can work quickly in real-time.”
“But, while the Viiola-Jones algorithm was something of a revelation for faces seen from the front, it cannot accurately spot faces from any other angle. And, that severely limits how it can be used for face search engines,” MIT wrote. “Which is why YAHOO is interested in this problem. This month, Sachin Farfade and Mohammad Saberian at YAHOO Labs in California; and, Li-Jia Li, at Stanford University nearby, reveal a new approach to the problem….that can spot faces at an angle, even when partially occluded. The two scientists say their approach is simpler than others; and yet, achieves state-of-the-art performance.”
“Farfade and Saberian use a fundamentally different approach to build their model,” known as ‘deep convolutional neural network. This technique involves training a computer to recognize elements of images from a database — using various layers,” according to Victoria Woollaston, writing in the February 18, 2015 DailyMailOnline. The two scientists created a database of 200K images, that included faces at various angles, and orientations; and, a further 20M images without faces. They then trained their neural net in batches of 128 images over 50K iterations,” MIT noted. “The result,” MIT contends, “is a single algorithm that can spot faces from a wide range of angles, even when partially occluded. And, it can spot many faces in the same image — with remarkable accuracy.”
The YAHOO team calls this approach, ‘Deep Dense Face Detector; and say, it compares well with other algorithms. “We’ve evaluated the proposed method with other deep learning based methods; and, showed that our method results in faster, and ,more accurate results ‘they say.’ “What’s more,” they add, “their algorithm is significantly better at spotting faces when upside down, something other approaches haven’t perfected. And, the researchers add, “it can be made even better…with data-sets that include more upside down faces. “We’re planning to use better sampling strategies; and more sophisticated data augmentation techniques…to further improve performance of the proposed method for detecting occluded, and rotated faces.”
Deep Dense Face Detector finds a match with 97.25 percent accuracy
MIT Technology Review says that YAHOO’s work “shows how fast facial recognition/detection is progressing. The deep, convolutional neural network technique is only a couple of years old itself; and, has already led to to major advances in object, and face recognition. The great promise of this kind of algorithm is in image search,” MIT Technology Review wrote. “At the moment, it is straightforward to hunt for images taken at a specific place, or at a certain time. But, it is hard to find images taken of specific people. This is a step in that direction; and, it is inevitable that this capability will be with us in the not too distant future. And, when it arrives, the world will become a much smaller place. It’s not just future pictures that will become searchable; but, the entire history of digitized images…including vast stores of video and CCTV footage. That’s going to be a powerful force, one way…or another,” MIT Technology Review concluded. But………….
Never Forgetting A Face: What Hath Facial Recognition Wrought
Natasha Singer had a lengthy article in the Sunday May 18, 2014 New York Times, “Never Forgetting A Face: What Hath Facial Recognition Wrought,” She begins by noting that Physicist Dr. Joseph Atick is a pioneer in the burgeoning field of “modern facial recognition. She adds, “the global business of biometrics — using people’s unique psychological characteristics, like fingerprint ridges and facial features, learn, or confirm their identity — is booming. It generated an estimated $7.2B in 2012,” according to the global consulting firm Frost and Sulivan.
“Face-matching today, could enable mass surveillance, basically robbing everyone of their anonymity,” said Dr. Atick. “Just a few months back,” Ms. Singer writes, “Dr. Atick heard about Name Tag, a [downloadable] app that, according to its news release, was available in early form…to people trying out Google Glass. Users had only to glance at a stranger and NameTag would instantly return a match — complete with that stranger’s name, occupation, public FaceBook profile information.” “We are basically allowing our fellow citizens to surveil us,” Dr. Atick remarked to Ms. Singer.
“Dr. Atick is just as bothered by what could be brewing quietly in larger companies,” writes Ms. Singer. “Over the past few years,” she adds, “several tech giants have acquired face-recognition start-up businesses. In 2012, Google bought Pittsburgh Pattern Recognition, a computer vision business — developed by researchers at Carnegie Mellon University. In 2013, FaceBook bought Face.com, and Israeli start-up.”
Dr. Atick says, “the technology he helped cultivate, requires some special safeguards. Unlike fingerprinting, or other biometric, or other biometric techniques, face recognition can be used at a distance…without people’s awareness; it could then link their faces and identities to the many pictures they have put online.” But, in the U.S., Ms. Singer writes, “no specific federal law governs face recognition.”
“As with many emerging technologies,” concludes Ms. Singer, “the arguments [with respect to facial recognition] tend to coalesce around two predictable poles: those who think the technology needs rules and regulation to prevent violations of civil liberties and, those who fear that regulation would stifle innovation. But, face recognition stands out,” she contends, “among such technologies; while people can disable Smartphoe geo-location and other tracking techniques, they can’t turn off their faces.” Or, can they?
Anti-Surveillance Mask Can Hide You From Biometric Face Scanners: “An Alternative Identity While In Public”
Mac Salvo had an on-line article in the May 13, 2014 SHTFplan.com, writing that, “just as quickly as governments introduce the technologies that are supposed to keep us safe from terrorists, and ourselves, enterprising rebels across the country are working [on ways] to counter them. This latest counter-surveillance technology comes in the biometrics/identity management domain.” “With literally hundreds of thousands of cameras now watching our every move. — and, plugging directly into [big] data-mining fusion centers, — where are activities are analyzed, aggregated, and dispatched according to our perceived threat, some might think the system is unbeatable. “Short of plastic surgery,” asks Mr. Salvo, “how can we modify our faces, to disappear from prying eyes — when we step out our front door?”
“If Leo Selvaggio has his way,” writes Mr. Salvo, “you’ll be able to assume an alternative identity by using age-old, low-tech strategy made possible by modern-day 3D printers.” “It is so simple it is brilliant,” adds Mr. Salvo, who contends that Mr. Selvaggio’s “innovation is capable of compromising multi-billion dollar facial recognition surveillance systems — with the use of an easily obtainable, personal prosthetic mask, or Personal Surveillance Identity Prosthetic (PSIP). It is one of three products made by the Chicago-based URME Surveillance, a venture dedicated to “protecting the public from surveillance; and, creating a safe space to explore our digital identities.”
“The 3-D printed, resin mask, made from a 3-D scan of Selvaggio’s face, and manufactured by ThatsMyFace.com, renders his features and skin tone — with surprising realism, though the eyes peeping out from the eye holes do lend a certain creepiness to the look,” noted Mr. Salvo. “When you wear these devices, the camera will track me instead of you and your actions in public space will be attributed as mine because it will make the cameras see,” the artist who’s working toward his Maters in Fine Arts at Chicago’s Columbia College. “All UMRE devices have been tested for facial recognition; and each probably identifies the wearer of me on FaceBook , which has some of the most sophisticated facial recognition around.”
Mr. Salvo notes that “the anti-face recognition technology is currently available in Leo Selvaggio’s image, so government systems spotting anyone wearing the mask will flag him as the culprit. But the implications are so broad that somewhere inside the Department of Homeland Security, surveillance personnel are undoubtedly scrambling to thwart it, because it presents a serious hiccough to the surveillance state With the ease of 3-D printing, any technophile with the ability to mimic someone else’s face via 3-D graphing software will have the ability to literally assume a person’s identity by simply printing their face and wearing it. In a world of biometric surveillance, that means anybody can disappear from view and essentially become a 21st century Silence Dogood.” (Mrs. Silence Dogood was the false persona used by Benjamin Franklin (at the age of 16 in 1722) to get his work published after the New England Courant had denied several previous attempts by the young Mr. Franklin to have his work appear in print).
Obviously, as this technology matures, it has the potential to be a major disruptor in the biometric/identity management domain — for both good and bad; and bridging well beyond the clandestine use of facial disguises historically utilized by the Intelligence, and Law Enforcement communities — among others. This development might even make the late, great actor Lon Chaney jealous. In the 1957 film noir classic, The Man Of A Thousand Faces, actor James Cagney played the title role which detailed the life of silent movie actor Lon Chaney. Mr. Chaney’s remarkable ability to transform his face and appearance — based on the character he was playing at the time, made him a legend on the movie set in the early days of Hollywood — with The Phantom of the Opera and The Hunchback of Notre Dame recognized as perhaps his best work.
Chris Smith, writing in the March 30, 2014 edition of TechRadar, “The Future Of Facial Recognition: Big Brother Or, Our New Best Friend,” writes, “that for better or worse, facial recognition has become the technological elephant in the room.” “Despite the continued advancements of software and algorithms — that would elicit gasps of awe in some of the other tech sectors, the ever-improving ability for machines to put a name to human faces is considered, in most cases, unwelcome.” “FaceBook’s recent unveiling of their DeepFace Research Paper,” he writes, discusses the “algorithm — which is still seemingly a long way from being integrated into consumer-facing element of the social network — that uses 3-D analysis of human faces to identify [a person of interest] — with a 97.25 percent success rate (and 99 percent under ideal circumstances); versus, our own human-brain/innate ability of an accuracy of 97.5 percent.”
Mr. Smith contends that “we’re on the precipice of a coming out party for the technology. Facial recognition has been simmering beneath the surface, reticent, and also unwilling to show its real face to the world, but it is now almost ready to emerge.”
But, as Lee Selvaggio has shown, if the Intelligence Community, Law Enforcement, etc., build a better mousetrap [facial recognition technology] enterprising rebels, malcontents, and others, will find a way to defeat, thwart, and/or, render these collection methods feckless in some circumstances. Once again, lots to think about. V/R, RCP