Tencent cloud open map technology accelerates intelligent security development

In the Hollywood blockbuster "Speed ​​7" there is a system called "Eye of the Sky". It can call a camera anywhere in the world, use face recognition technology to search for the person or thing you want, and let it be nothing. In stark contrast to this, when it comes to real-life security, it still relies on the reports of the Chaoyang people to crack down on illegal behavior. The netizen said: "The Chaoyang people have become world-class intelligence agencies that can be named in the same way as FBI, KGB, and MI6."

Behind the ridicule reveals a serious shortcoming of intelligence in the security field, and the face-shaping technology that is booming at the current stage opens a window for the breakthrough of intelligent security . Recently, Tencent Cloud announced the opening of the excellent face recognition technology at the first Technology Leaders Summit. When face recognition and cloud computing are combined, it is bound to accelerate the intelligentization and de-artification of the security field.

Excellent picture technology

Face recognition vs. Chaoyang masses

China has a vast territory, a large population, and a strong criminality. A variety of factors have made the security situation more complicated, and it is necessary to carry out accurate identification and information identification for large-scale population through technical means. Therefore, adopting intelligent security measures has important significance in the fields of national security, public security, justice, e-commerce and security systems.

Based on this, many smart city managers have turned their attention to face recognition. At the moment, face recognition can already achieve a reduced version of the "Eye of the Eyes" system, installing face recognition systems in crowded areas such as train stations, airports, and subway stations. It can effectively identify key targets such as suspects and dangerous goods.

But even then, it still does not solve all the problems. The first is that the accuracy of face recognition is not high. Although digitally, many face recognition technologies can achieve an accuracy rate of more than 90%, but the security problem is different from other problems, and a small omission can cause great losses. Therefore, it is necessary to ensure that the accuracy rate is infinitely close to 100%, so that the security responsibility can be safely given to face recognition.

Second, the key to face recognition technology, in addition to identification, is the inclusion of massive data comparisons. This puts high demands on the server capacity in the background. In order to achieve high-speed and accurate face recognition, it is necessary to realize retrieval and calculation in 10 million or even hundreds of millions of photos, and to extract important information from it.

Face recognition needs to be combined with cloud computing

In response to the above problems, Tencent Cloud promoted the full openness of Uto face recognition technology and combined face recognition technology with cloud computing and big data. Huang Feiyue, the head of Tencent's excellent map team, said in an interview: "To apply face recognition technology in actual business, there must be a considerable amount of training data consistent with the actual application scenario, in order to ensure face recognition algorithm. The effect is good enough in this application scenario."

In this respect, Tencent's face recognition has a unique advantage. In terms of quantity, Uto has a billion-dollar big data processing platform. In the QQ space photo album platform, the U-Map team handles hundreds of millions of photos in a single day, and has accumulated tens of billions of photos. In terms of accuracy, the U-Map team has a very large-scale training set, and the accuracy of face recognition is currently ranked first in the world. Not long ago, on the international authoritative face recognition database LFW, the Tencent Youtu team submitted the latest score of 99.65% in the face verification test under unrestricted conditions, creating a new record and reaching the world's first level.

Combined with the powerful computing power of Tencent Cloud, Uto face recognition can quickly complete face comparison, verification and recognition in massive data. For example, comparing the similarity of faces in a given two images and the similarity of facial features; determining whether a given face image is the same person and degree to a certain face; and detecting a given Face image and retrieve his identity information from the database and return.

Excellent picture face recognition on Tencent Cloud

The UTU team currently uses OpenAPI to open up the core capabilities of face detection and face recognition. It provides 11 core APIs for face detection, face registration, and face comparison for users to use, which can realize face detection. 1:1 face verification, 1:N face search and other functions, developers can use this package to land a variety of application scenarios.

The Tencent U-Map open platform was officially released in June, and core technologies including face detection and face recognition began to open to the public. The opening of U-Map on Tencent Cloud will further promote the widespread reference of face recognition technology in various industries. Developers with Tencent cloud accounts can directly invoke the face recognition cloud service after login, without application review. Such an action is undoubtedly not a big pie falling from the sky for the security field introduced by the face recognition technology.

Face recognition technology can bring subversive intelligent changes to the crimes of crime prevention, protection of urban security, intelligent access to public places such as campuses, office places, airports, etc., and even smart home locks in houses and vehicles in family life. . The continuous opening of Tencent's technology, especially in combination with Tencent's cloud video solutions, will promote the widespread use of face recognition technology in the security field, bringing more stability to a comfortable city and human comfort production and living environment. Security.