Facebook 100 million dollar buy face: buying face.com behind

Sch Jian Guangzhou reported that Facebook was originally called the Facebook.　　Some abuse: Now, it is not want to take the book also removed, directly called face? June 18--The face of an Israeli-backed facial recognition venture company. COM announced that it was bought by Facebook, which has been heavily invested in the IPO.　　The two sides did not disclose the amount of the deal, according to analysts here predicted that Facebook for the "face", spent 80 million-100 million dollars.　　If Facebook is going to spend so much money on just one domain name, it's not a joke, it's a lack of awareness of the rising face recognition Internet application. Did Facebook buy two identical leaves in the world? This is a question that philosophers think about.　　And you can't find another face like yours, it's the theoretical foundation of facial recognition technology. Human face and fingerprint, iris and so on have the uniqueness, so that it has the detection, recognition and other aspects of the application value.　　In fact, face recognition is not a new technology, as early as 40 years ago, there have been technical experts began to delve into this, its application in all walks of life in-depth, such as security, access control attendance, anti-theft door, and even spy industry. For example, in the upcoming movie "Mission 4", the agent is using contact lenses to scan the face of the search target, and at the railway station to intercept the secret files.　　While in the alley, when he was killed by the oncoming female killer, the face recognition lens in his eye had photographed the killer's facial features and entered his cell phone, which almost simultaneously resolved the killer's picture information. The Beijing Olympic Games has also introduced the face recognition technology in the security aspect, has realized the Olympic Games competition to carry on the real name identification and the identity identification function to the ticket holder.　　In the upcoming London Olympics on July 28, it is said that face recognition will come in handy again, and visitors to the venue, in addition to the card verification, also need to collect human facial information from the camera scene to compare to more accurate identification.　　Although face recognition has accumulated more and more sophisticated technology in many industries, its large-scale applications on the Internet have to be the top of the Facebook platform. In fact, Facebook has worked with the start-up company for nearly two years before acquiring Face.com.　　Facebook began a small scale test of face recognition in July 2010, with the aim of simplifying the user's ability to identify friends on photos. Currently, the number of photos uploaded to Facebook has reached 300 million a day. With face recognition technology, Facebook can automatically identify people who upload photos and then prompt users to label them.　　Before this, users can only manually complete the action to tag friends photos, for users with a large number of photos, this operation is quite cumbersome. FAC, which provides Facebook with photographic facial detection and recognition techniquesE.com, a start-up that was founded in Israel. "Our mission is and always has been to find new and exciting ways to make facial recognition technology interesting, to make it a part of people's lives, to combine superior technology with everyday consumer goods."　　"June 18, face.com in a blog announcing the company's Facebook takeover," he wrote.　　In addition to developing third-party apps used on Facebook platforms photo tagger, Face.com has a celebrityfindr that is used on Twitter platforms to help users quickly detect photos that celebrities want to make public. Face.com is not satisfied with its own development applications, it on the Web page for its own banner is, "To do the world's largest and most accurate face recognition platform." For this, face.　　COM opens APIs (application programming interfaces) to other application developers, allowing Third-party developers to use their face recognition techniques in their applications.　　Face.com said that even after being bought by Facebook, it would not stop API open support for Third-party application developers. Precision and user experience of the balance beam "face recognition technology has been in existence for decades, and its recent years in the Internet application of the emerging, largely due to the popularity of the camera."　　July 6, NetEase Hangzhou Research and Development Center, the head of the face recognition program told reporters that the popularity of the hardware is the face recognition before the iris recognition, handprint recognition in the Internet application of an important factor. In terms of hardware, with the advent of applications such as video calls, the camera has become the standard for laptops, and the proportion of desktop users buying separate cameras is also high. As for mobile phones, cameras are not only standard, but the camera pixel has become the handset manufacturers to attract users of important indicators, from 300,000, all the way to 2 million, 5 million, 8 million.　　Even Nokia, which launched a camera early this year, has made the camera's pixel 41 million.　　In contrast, iris recognition, handprint recognition, although the accuracy is higher, but ordinary Internet users also difficult to equip with special identification equipment. According to the NetEase person in charge, NetEase's face recognition technology application, first put on the security of its mailbox account. NetEase mailbox in the Chinese market has the largest user base, has been, NetEase CEO Ding Lei on the business market position also very mind. Up to now, NetEase has launched a general order, secret security card, mobile phone secret security, such as a series of password protection measures.　　And at the end of last year Csdn (Weibo) and other websites burst account of large-scale leakage accident, NetEase accelerated the face Recognition technology application research and development. NetEase's face recognition project began in 2010, it is said that the team has more than Dr. and returnee talent.　　After nearly two years of brewing, this May 30, NetEase in the mailbox users launched a Face recognition trial user recruitment.The reporter obtains the data from the NetEase company, the recruitment process, daily applies for the trial user average more than 20,000 people, netease every day by the system from the day registers the user randomly to collect more than 200 users, up to now a total of more than 7,000 users participated in netease face recognition trial. In the laboratory before the user trial stage, NetEase said its facial recognition technology accuracy can reach 99.9%.　　But in the actual trial process, due to the user-side camera pixel, ambient light, user attitude and other factors, the recognition accuracy will be reduced. "From the current point of view, face recognition can only be used as an auxiliary means of NetEase mailbox security authentication."　　The head of NetEase face recognition project mentioned above said. She further said that, technically speaking, accuracy can be higher, but as an Internet application, NetEase needs to balance between accuracy and user experience, "not the technology is not achieved, but to the unilateral pursuit of accuracy, may be more users to ask,　　It may also give users more time to wait for authentication, which obviously undermines the user experience most valued by the Internet business. The NetEase person cited two examples.　　The first is the twins, there are already research institutions developed to identify the height of the same characteristics of the twins, but the application cost is high, such as the need for a very high definition of the camera, the algorithm to achieve the difficulty and performance of the Internet is not suitable for large-scale application of massive data. Another is the photo, if the user in the use of other photos to log in to the mailbox, whether the smooth passage? To this, the NetEase's answer is that for the distinction between photos and real people, NetEase has related algorithms to support, its distinction between photos and real person accuracy is higher than the current industry has published academic papers. When face recognition is officially released, this feature will be added as an upgrade feature, usually avoiding the ability to log in with photos.　　In practical applications, due to the influence of complex factors, the accuracy rate may be slightly lower than the experimental environment. Privacy concerns In addition to NetEase, Tencent, Grand Innovation Institute, etc. also in the face recognition of the action. Grand Innovation Institute of a mobile phone photo management software-Smart album, will be applied to the grand upcoming Bambook smartphone.　　It can classify photos through face recognition technology, improve the efficiency of photo retrieval in photo albums, and support direct phone calls from photos. Tencent Institute has also developed its own face authentication login system. Based on this, Tencent also launched a face recognition of the entertainment application-search eye. For example, users can take photos of themselves, and then use this app to compare with the photos of the stars, and upload the results to the microblog to share with friends.　　Of course, this application is relatively less stringent in terms of identifying technical accuracy, and has been a purely entertaining category. NetEase face recognition related to the reporter said that the current NetEase face recognition technology applications are mainly in the mailbox-assisted security certification, the future does not rule out its embedding more Internet products, such as NetEase games, and even do not excludePossible for a platform launch.　　But this premise, in addition to solving the actual application process accuracy problem, but also must pay attention to another factor-the protection of user privacy. "For functional applications such as mailbox security authentication, because it's one-to-one and without privacy issues, it's complicated to make Internet products, and Facebook has suffered a loss in this area," he said.　　"Some people in the industry said.　　Last June, privacy agencies such as the Electronic Privacy Information Center, consumer protection organizations, and privacy-exchange centers complained to the Federal Trade Commission (FTC) that the FTC would force Facebook to stop the biometric information of mobile users without the consent of the user. Given this lesson, Google +, when it comes to a similar face recognition application--looking for my face--will give users the freedom to choose whether to use the service automatically before exposing the photos, and only if Google + users agree, will the system give the user face recognition.