A team of researchers from the University of Texas at Austin and Cornell Tech have trained a software program to uncloak digitally-blurred or distorted images using deep learning, essentially teaching a computer to interpret a set of example data. Using this process, the team's software identified encrypted photographs with a 71% success rate. For context, the human success rate was 0.2%. The team purposely used an open source deep learning library to train their software, and their research exposes more weaknesses in the concept of online privacy. "The techniques we're using in this paper are very standard in image recognition, which is a disturbing thought," said Cornell Tech's Vitaly Shmatikov, pointing out that theirs was an "off-the-shelf, poor man's approach" to encrypted image recognition, and that a person or entity with bad intentions could do a lot of damage with this technology.

It's becoming much easier to crack internet privacy measures, especially blurred or pixelated images. Those methods make it tough for people to see sensitive information such as obscured license plate numbers or censored faces, but researchers from University of Texas at Austin and Cornell University say that the practice is wildly insecure in the age of machine learning. Using simple deep learning tools, the three-person team was able identify obfuscated faces and numbers with alarming accuracy. On an industry standard dataset where humans had 0.19% chance of identifying a face, the algorithm had 71% accuracy (or 83% if allowed to guess five times). The algorithm doesn't produce a deblurred image--it simply identifies what it sees in the obscured photo, based on information it already knows.

Pixelation is used heavily across the media to obscure faces, vehicle license plates and restricted pieces of text. The technique generally works well, preventing people from identifying what's beneath the blur without completely destroying the image with overlays or cut-outs. However, we may soon need to find an alternative to pixelation, according to researchers at the University of Texas. WIRED reports that a team has developed software that is able to see through a pixelated mask to determine the identity of a person beneath it. The technology is able to process images as if the pixelation simply didn't exist.

A joint UT Austin/Cornell team has taught a machine learning system based on the free/open Torch library to correctly guess the content of pixellated or blurred redactions with high accuracy: for masked faces that humans correctly guess 0.19% of the time, the system can make a correct guess 83% of the time, when given five tries. Redaction errors have plagued data-releases since the earliest days of the net; who can forget the hilarity of companies and agencies that added black boxes in an overlay to their PDFs, or left Word's document history (including all the deleted passages) intact on their sensitive releases? Or the pedophile whose twirly-faced redaction was de-twirled to catch and prosecute him? These days, the best practice seems to be opening the images in a bitmap editor, then replacing them with black squares. "We're using this off-the-shelf, poor man's approach," says Vitaly Shmatikov, co-author of the paper and professor at Cornell.

Believe it or not, but a recent study conducted by researchers at the University of Texas and Cornell University has revealed that [PDF] the blurring technology is soon to become exploitable if not obsolete. We may not enjoy full confidentiality that we do now by blurring pictures or license plates for much long because now computer devices can decode images using Artificial Intelligence. This means it is quite easy to unblur pixelated or blurred pictures using various readily and easily available software tools that help in identifying faces or information. The team of researchers utilized a range of deep learning tools to unblur 71% of the blurred faces and numbers, while the percentage increased to 80% when they allowed the computer to guess for up to 5 times. It is true that their algorithm cannot create the original image but it can identify whatever can be seen in the blurred photograph according to the information that it has acquired.

Researchers fed the software this picture of actor J.K. Simmons, among others, to teach it to recognize specific faces. Since 1989, Cops has famously aired footage of suspected criminals, many with their faces blurred out to protect their privacy. Ever since then, blurred or pixelated faces have become standard fare for concealing the identity of individuals who prefer not to be recognized in the media. YouTube got in the game a few years ago, offering a facial blurring tool to help protect protestors against retribution from law enforcement or employers. But machine learning researchers at Cornell Tech and the University of Texas at Austin have developed software that makes it possible for users to recognize a person's concealed face in photographs or videos.

Blurring or pixelating information to obscure it may not work anymore thanks to machine learning researchers from the University of Texas at Austin and Cornell University. The researchers developed an algorithm that could identify faces and numbers even after they were blurred out. The researchers developed the algorithm using open-source machine-learning software. Just take a bunch of training data, throw some neural networks on it, throw standard image recognition algorithms on it, and even with this approach…we can obtain pretty good results." The algorithm is built using a very simple process.

Pixelation has long been a familiar fig leaf to cover our visual media's most private parts. Blurred chunks of text or obscured faces and license plates show up on the news, in redacted documents, and online. The technique is nothing fancy, but it has worked well enough, because people can't see or read through the distortion. The problem, however, is that humans aren't the only image recognition masters around anymore. As computer vision becomes increasingly robust, it's starting to see things we can't.

The recent emergence of a powerful machine-learning technique known as deep learning has made computing giants such as Google, Facebook, and Microsoft even hungrier for data. It's what lets software learn to do things like recognize images or understand language. Yet many problems where deep learning could be most valuable involve data that is hard to come by or is held by organizations that are unwilling to share it. And as Apple CEO Tim Cook puts it, some consumers are already concerned about companies "gobbling up" their personal information. "A lot of people who hold sensitive data sets like medical images are just not going to share them for legal and regulatory concerns," says Vitaly Shmatikov, a professor at Cornell Tech who studies privacy.

Technology giants such as Google, Facebook Inc (NASDAQ:FB) and Microsoft Corporation (NASDAQ:MSFT) have been made hungry for more data thanks to the recent emergence of powerful machine learning techniques known as deep learning in Artificial Intelligence. Using Artificial Intelligence or Deep learning is what gives software's ability to learn to do things such as recognizing images and understanding languages. Many of the problems that deep learning could be an invaluable asset, however, involve times when data is hard to come by or in some cases held by organizations which are unwilling to share it. According to Apple CEO, Tim Cook, some customers are already concerned about companies which take up all of their private and personal information. Vitaly Shmatikov, a professor at Cornell Tech who studies privacy says, "A lot of people who hold sensitive data sets like medical images are just not going to share them for legal and regulatory concerns.