In the security industry, the real technology in the past 20 years has focused on the front end, not the back end, and these years have gradually shifted to the intelligent direction.
The author: Synopsys CEO Ma Yuan Peng
Title of this article: “ Industry Opportunities and Reconstruction of Intelligent Views in the Age of Late-Launch Advantages of AI”
Editor: Shi Yaqiong
In 2018, China’s overall security industry output value was about 709.6 billion yuan, and the market size is expected to reach 741.2 billion in 2019. As one of the fastest-growing and fastest-growing markets for artificial intelligence technology, the market size of AI security is expected to be 35 billion yuan in 2019, and AI penetration is only 6%. It is estimated that by 2020, the size of the AI + security software and hardware market will reach 45.3 billion yuan.
After the development in recent years, AI security has only opened the landing application of face recognition technology. Facing such a huge smart view blue ocean market, there will be more technical development space and landing scenarios in the next decade worthy of attention. At present, the evolutionary stage of technology has moved from the face-oriented AI security 1.0 to the full-scene analysis of AI security 2.0. The security industry view application opens the reconfiguration of “function machine to smart machine”, and the evolution of security and mobile phone applications will experience hardware- and performance-driven to software and data-driven.
From the perspective of technology and market, future AI security has room for rapid growth. In the technical dimension, the AI algorithm will expand horizontally from the analysis of the face to the full elements of vehicles, pedestrian ReID, behavior, and scenes; in the market dimension, it will penetrate vertically from security to industry scenarios such as communities, campuses, buildings, and parks. .
In the security industry, the real technology in the past 20 years has focused on the front end, not the back end, and these years have gradually shifted to the intelligent direction. So, in fact, for the time being, everyone is facing the same starting line: the entire industry has begun to transition to technology, data, and services.
In the stage of AI implementation, it is no longer the “gold rush era” of early AI + industry companies. Using AI technology to empower industries, it has entered the “alchemy era” of industry + AI companies. This type of company has its own industry attributes, continues to cultivate in an industry, and finally solves the problems facing the industry.
In a market that is strongly related to scene data, a vertical industry + AI company will inevitably appear. The AI security industry is currently a vertical field that is very suitable for “alchemy”. As a typical vertical industry, security is all image and video data, which cannot be solved with general-purpose technology. It must be done with vertical, dedicated image recognition technology, such as multi-angle and multi-pose, face recognition under complex light. When Shenzha entered the security industry, general AI technology began to split. In the early stage, a general technology was considered acceptable. By the second stage, general technologyThe technology cannot meet industry requirements, and it must be AI technology with industry attributes.
In the security industry, a company like Hikvision 2.0 or Dahua 2.0 is bound to emerge, just like the process of the mobile phone market from Nokia and Motorola to Apple, OPPO, vivo, and Xiaomi. The final form of AI security needs to be closed-loop with the value of “AI front-end + SaaS + cloud”. The core point of an enterprise is that it must have full-scale algorithm capabilities, but also have “algorithm + IoT + software + cloud” full stack solution capabilities .
This is different from the Internet industry. It is not just a matter of doing a good job. Instead, it is necessary to open up the entire industry chain to cover the demands of the entire security industry.
Each technological innovation will undergo a long evolution process. As an application technology, artificial intelligence technology is also undergoing such an evolutionary process. Initially, the technology can only solve a certain problem in a specific field, but as general technology goes from popularization to the industry, it will face more and more segmented scenarios, which requires the combination of technology and scenarios to target different industries Emerging issues optimize technology to suit and meet specific needs.
After having the full stack of technical algorithm capabilities, enterprises also need to have intelligent hardware products for all scenarios. At present, everyone has different approaches, including cooperation, empowerment, and authorization. PENS Technology adheres to the full self-developed route of hardware. For the segmented scenes of image acquisition, edge intelligence, security inspection, identity verification, personnel access, welcome attendance sign-in, temporary security, etc., it introduces intelligent face recognition access control and intelligent face analysis box. , Gate companion, intelligent portable portrait control system, intelligent portrait snap camera, dome network camera and other hardware products.
Looking forward to the next time, while AI technology is developing, landing will become a more important keyword. The AI industry has begun to compete from technological innovation to business models and landing strength. The evaluation criteria of AI companies should gradually change from financing scale, winning championships, and technological empowerment to deep cultivation of vertical scenarios, extreme product experience and innovative application services. In the era of late-generation advantages of AI landing, Pengsi Technology, as the second-stage AI company, adheres to the “industry + AI” strategy, deeply cultivates vertical industries, and becomes a product-based company.