On November 7th, the "Huawei Fully Connected Conference 2022" was held in Shenzhen, along with the Huawei Urban Intelligent Agent Summit. At the summit, the China Institute of Electronic Technology Standardization, Huawei Technology Co., Ltd., and Minivision Technology (Jiangsu) Co., Ltd. jointly released the "White Paper on Urban Perception System (2022)". In the field of urban perception, many mature projects of Minivision Technology have been implemented, such as fire monitoring, high-altitude parabolic monitoring, septic tank monitoring Air quality monitoring and other systems have also been applied in scenarios such as smart communities.
The progress of modern technology has brought us into an era of interconnected things. As the first stop for the development of social civilization, our city is also moving towards intelligence and digitization. The report of the 20th National Congress of the Communist Party of China pointed out that "we need to strengthen urban infrastructure construction and create livable, resilient, and smart cities." It can be seen that the digital transformation of cities has become a key measure to implement the spirit of the 20th National Congress of the Communist Party of China and implement the strategy of digital China and regional coordinated development.
The importance of building a city perception system lies in its ability to break the barrier between the physical world and the digital world, shape the root "nerve endings" of the city with technology, and create a smart city that combines visual, auditory, tactile, and olfactory senses, thereby mastering the myriad scenarios of urban operation. Among them, Xiaoshi Technology focuses on how to build visual perception capabilities for cities based on AI technology, or how to activate a city's video surveillance network and build a standard, unified, and open visual center.
Based on core technologies such as computer vision algorithms, video AI analysis, and big data, Minivision Intelligent Vision Hub Platform can perceive the massive data of urban video surveillance, while strengthening the utility of "data empowerment" in urban refined perception and governance.
By relying on the "small sample weak supervision technology", the algorithm production efficiency of Minivision Intelligent Vision Central Platform can be increased by more than 6 times, and the speed of algorithm iteration and accuracy improvement can also be significantly improved. This means that the platform can provide diversified algorithm customization services, and the long tail of urban governance needs to be met. The scheduling strategies such as video rotation, tidal scheduling, and resource pooling have increased the utilization rate of computing resources by more than 50%, allowing the platform to access millions of videos and undertake city level video analysis tasks.
Building an intelligent visual center is an important step taken by Minivision Technology in fields such as smart cities and urban visual perception. In the future, Minivision Technology will actively participate in the construction of a smart city perception system and standards based on this, and fully promote the integration of urban digital transformation blocks.
