In recent years, after the wide application of 3D, VR, AR and other technologies, artificial intelligence has become a new outlet. All of a sudden, the capital of AI products are in great demand, all industries hope to work with AI, security field is no exception. In fact, with the continuous development of the economy, the continuous maturity of AI technology, coupled with the attention of national policy, the size of the domestic security industry is also growing. According to Jiadu Technology “Artificial Intelligence Technology white” shows that the size of the security market in 2017 will rise from the previous forecast of 614.7 billion to more than 635 billion, a year-on-year growth of 17.6%, the number of security manufacturers more than 7,000.
In the age of intelligence, AI security is coming
With the maturity of AI technology, intelligence has become the new development direction of the security industry. The National Development and Reform Commission, Ministry of Science and Technology, Ministry of Industry and Information Technology, the central net letter do make the “Internet” action plan for three years of artificial intelligence, encourage security cooperation between enterprises and the Internet, research and development of integrated image and video identify precisely, biometric recognition, coding and so on the many kinds of technology of intelligent security products, Promote the intelligence, intensification and networking of security products.
Now, intelligent monitoring systems have covered many applications, so it can be seen that AI security will replace traditional security technology in the near future. In front of the huge market prospect, security enterprises have held high the FLAG of AI, in order to seize this rare market reform opportunities.
It is reported that the current security field has haikang and Dahua two giants, in 2017, Haikang and Dahua two will occupy 43% of the market share, revenue of more than 10 billion, net profit of more than 1 billion. When other entrepreneurs are still exploring the way of AI security, the two giants have taken a step ahead of others to carry out the strategic deployment of “AI + security”.
In 2015, Haikang launched “Falcon” and “Blade” structured servers based on AI deep learning technology. In 2016, it released a series of AI products such as “Deep Eye” camera, “Super Brain” NVR, and “Facebook” face analysis server. In addition, Hikon has formed partnerships with Nvidia, the largest maker of GPU chips for ARTIFICIAL intelligence, and Movidius, a dark dog for computer vision chips, to continue to dig deeper into the AI security industry.
Coincidentally, dahua, another security giant, began to explore AI security in 2015. In 2016, it established four research institutes of chip, Big Data, artificial Intelligence and Advanced Application, developed front and back end products such as face recognition, video structuring, abnormal behavior analysis and high-density crowd analysis based on deep learning technology, and proposed corresponding solutions. In 2017, DAhua released a number of “wise” series of video structured servers.
Sensetime, a newcomer in the field of AI security founded in 2014, completed a $410 million B round of financing in July 2017. Its core technologies include face technology, intelligent monitoring, image recognition, and text recognition. As an AI startup, the deep professional talent pool is the fundamental reason that makes Sensetime stand out in the field of machine vision technology. At present, Sensetime has 18 professors and more than 120 doctoral students from world famous universities. These academic networks composed of colleges, teachers and students have been contributing talents in related fields to Sensetime.
At present, the main products of SenseTime include SenseID authentication solution, SenseFace face control system, SensePhoto mobile phone image processing solution, SenseGo smart business solution and SenseAR augmented reality special effects engine, etc. It can be seen that Sensetime’s layout of video images is deeply intelligent, and the intention of deep cultivation of AI security is very obvious. The AI war in the field of security has begun…
AI security is not an easy nut to crack
In fact, AI security is not a good bone to chew, in AI security novices to march at the same time, to Haikang and Dahua represented by the rich and powerful old security companies, are also increasing the investment in AI technology. In the major enterprises are sparing no effort to find the combination of security and AI at the same time, with a variety of AI-related security products, technologies and programs gradually increasing, the emergence of more and more problems.
First of all, the cost of intelligent security is high, and Dahua, Haikang and other old security companies actively layout AI security, market competition pressure. Security industry has long been regarded as the AI technology to the ground the best industry, security industry if you want the AI technology universal coverage, AI enterprises are in urgent need of research and development of image sensor, camera module, and so on core parts and components, as well as the connection of upstream and downstream manufacturers for the contestant, and these components research and development and product promotion is bound to lead to higher costs.
For Sensetime, it is dominated by technical teams and lacks profound industry accumulation and industry resources. In the process of developing AI security products, sensetime’s technical team needs not only a long running-in time but also a large amount of financial support to develop products or propose solutions. Moreover, it is well known that AI engineers are generally paid very high salaries, with a basic monthly salary of more than 20,000 yuan, and the financing for start-ups is not enough to support the high cost burden brought by long-term research and development.
Secondly, for security novices, the key to the success or failure of AI security is the landing effect and whether it can continue to innovate and develop. Many consumers still have doubts about the functions of AI products. The reason is that most AI products in the market are regarded as showy tools, which are either over-advertised or technically unstable and cannot bring convenience to consumers. Therefore, before the successful combination of technology and products, the AI products made have not solved the problems of the industry itself, and thus will not be accepted by consumers. Therefore, there are still many problems to be solved in the application of artificial intelligence in the field of security.
In terms of the application of AI technology, Thomson technology platform, although the self-built algorithm algorithm itself does not solve any practical problems, but AI AI algorithms fall to the ground with a large number of application scenario for the material as a support, and many start-ups could not have this ability, so, Thomson is the key of the development of science and technology breakthrough technical difficulties as well as the application scenario of material collection, So as to create a good AI application program, to achieve innovative development.
In addition, at present, most AI security products are faced with practical problems that users do not accept and technology is not mature. Although the pursuit of intelligent security is almost a common goal, it does not mean that the market can easily accept the application of AI security. Although AI can indeed bring more development opportunities for the security industry, the market has been very rational in the choice of AI products. Therefore, for AI security enterprises, how to implement AI technology to users can use affordable and feel easy to use security products, security enterprises at the present stage to complete.
In general, Sensetime’s advantages mainly lie in the accumulation of talents and technologies in the field of computer vision, while its disadvantages lie in the lack of brand awareness, scene application and sample data. According to data disclosed by MarketLine, the global security service industry has shown a rapid development trend in recent years, but the growth rate is expected to slow down slightly, with a compound annual growth rate of 7.6%. And the arrival of the era of artificial intelligence, security industry will usher in a big outbreak, the competition between the major security enterprises will be more and more intense.
The landing of AI security products is the key for enterprises to stand out
According to the statistics of China Security Network, the security market size will reach 752.1 billion yuan in 2018. In addition, in recent years, biometric identification, big data, image processing, deep learning and other artificial intelligence technology research and development in the security industry can be described as the peak of the day, so far security products intelligent has become the development direction of the industry. Thanks to the development of data, algorithms, computing power and government support, the momentum of AI security development has been unstoppable.
The arrival of the intelligent era has injected new vitality into various industries and brought many opportunities to enterprises. However, there are also many competitors, and it is difficult to stand out from many enterprises. Artificial intelligence technology is iterated too fast. Although the research and development of artificial intelligence technology has taken root and developed in the security industry, the landing application of AI products is still an important link in the current development of ARTIFICIAL intelligence. Therefore, Sensetime can occupy a favorable position in the field of AI security only by constantly testing its technology in actual combat landing.
On the one hand, the growth of the security industry slows down, the market tends to saturate, the major security manufacturers product homogenization, increasingly fierce competition, urgent need to find new development opportunities, and the arrival of AI just provides such an opportunity. As security continues to move toward the direction of intelligence, AI will be fully integrated into all aspects of the security field in the future, leading security into the era of comprehensive intelligence.
On the other hand, most of the AI solutions in the security industry are to optimize the monitoring system and achieve the purpose of intelligent security through accurate analysis and prediction of data. However, as far as the current security situation is concerned, the material data is relatively lacking. Due to the lack of data, many predictions are not accurate. Therefore, AI enterprises are still facing the problem of AI security products landing, and there are certain difficulties in product promotion and application.
Overall, AI has been explored in a wide range of applications in various industries in recent years, and has tasted the benefits. In the field of security, with a large number of security enterprises and newly established AI enterprises began to work deeply, AI security products are increasing. At present, face recognition has been used in police surveillance, train and airport entry and exit stations and even in university classrooms. Other ARTIFICIAL intelligence technologies have also been well developed. However, in addition to the technical limitations of AI products themselves, the lack of data and high costs in the security industry limit the further development of AI in the security industry. Under such circumstances, AI companies can stand out only if they come up with products that people believe in.
Article/Liu Kuang public account, ID: Liukuang110