06,04,2026 22 views
The World Artificial Intelligence Conference (WAIC 2026) will be held in Shanghai from July 17th to 20th. Hikvision will make its debut at the conference, showcasing its Guanlan large model technology system in all aspects. It is reported that the Hikvision exhibition area will focus on how big model technology is deeply integrated into various software and hardware products and quickly applied; The big model is for industrial and urban governanceIntelligent TransportationThrough a wide range of scenarios, we aim to improve quality, reduce costs, increase efficiency, and ensure safety, becoming a true assistant for the intelligent upgrading of various industries.
Highlight 1: Born to serve thousands of industries, Guanlan's large model technology system is unveiled
Artificial intelligence technology has always been one of the core technologies of Hikvision, and an intelligent algorithm team was established as early as 2006. In the wave of booming development of big model technology, Hikvision is accelerating breakthroughs in model capabilities, productization, and deployment. It is not limited to polishing basic big models, but also focuses on building AI comprehensive capabilities that support high cost-effectiveness and large-scale implementation of big models.
Based on the Hikvision Guanlan big model, Hikvision has built an industry-specific big model technology system, covering "basic big models, vertical big models, big model products, and industry applications". It is committed to using strong IoT perception and cognitive capabilities to help better connect the physical and digital worlds, insight into the state and laws of all things, and promote the intelligent development of society, industry, and life.
In the application of large-scale models, the Guanlan large-scale model technology system has the systematic advantages of "multiple, fast, accurate, and economical". At the Hikvision exhibition area, you can fully understand the Guanlan large model technology system and see how AI can help understand the physical world and enter various industries.
Highlight 2: Showcase a series of large model software and hardware products, and release multiple new large model products
In recent years, by deeply integrating big model technology into products, Hikvision has released thousands of big model software and hardware products, achieving full coverage of cloud edge end. At the Hikvision exhibition area, a series of large model software and hardware products will be showcased on site, and new large model products such as multimodal intelligent cameras, active visual job supervision guns and balls, AI multimodal safety production hosts, and SOP training and push all-in-one machines will be released.
Multimodal intelligent camera: Deploying multimodal models of graphics and visual images into the camera not only enables understanding of the scene, but also enables the camera to automatically work with a single sentence, leading the camera from "understanding" to a new era of "understanding, dialogue, and work", from passive recording to active analysis and execution.
Active visual homework supervision gun ball: With the enhancement of visual modeling capabilities, cameras are no longer just passive recording, but can truly "understand the scene". They can actively patrol, discover safety hazards, and act as "digital safety officers" in the safety production supervision of enterprises.
Guanlan encoded camera and Guanlan encoded disk array: Integrating Guanlan large model and pixel level segmentation technology, providing high bit rate guarantee for high-value key targets (such as people, cars, and non motorized vehicles), efficiently compressing non critical areas, achieving differentiated encoding, saving more than 50% of bit rate, effectively saving storage space, reducing power loss, and solving the "storage dilemma" caused by high-definition monitoring.
An intelligent agent that can work with just one sentence: ChatBI - Data Analysis Intelligent Agent realizes "as long as you can type, you can conduct data analysis and judgment", for example, inputting "help me analyze the energy consumption situation of various communities and generate visual dashboards" to view the analysis effect; Cloud Eye CLAW has become the "digital employee" of chain management, capable of automatically "running" the entire process of store inspections and reviews with just one sentence; The emergency command intelligent agent can adjust videos, create meetings, identify hidden dangers, check resources, issue instructions, make phone calls, etc. with just one sentence, promoting simpler, more agile, and more realistic emergency command.
Highlight 3: Showcase the application of a series of vertical models and see AI truly penetrate industry scenarios
Industry scenarios have the characteristics of strong task specialization and high precision requirements, and general large-scale models are difficult to directly meet the intelligence needs of each segmented field. Hikvision is based on basic big models and deeply injects industry knowledge and business know-how into specific fields, forming a series of vertical big models covering safety production, industrial manufacturing, chain inspection, pan park, public safety, urban governance, traffic management, natural disaster monitoring, infrastructure inspection and other fields, achieving the optimization and adaptation of the general capabilities of big models to industry scenarios.
At the Hikvision exhibition area, you can intuitively feel the landing value of vertical large models: safety production large models help with special operation control, intelligent monitoring of leakage, and intelligent monitoring of belt conveyors; Industrial large-scale models assist in SOP testing, industrial quality inspection, and industrial temperature measurement for production line operations; The urban governance model covers urban waterlogging monitoring, third-party construction damage monitoring, and emergency command and dispatch, making urban operations safer and more orderly; The traffic management model accurately detects traffic incidents and intelligently regulates traffic signals, making travel smoother and management more efficient.
Highlight 4: Interactive Experience, Endless Excitement
The exhibition area has specially set up multiple interactive experience zones, allowing visitors to personally experience the intelligent changes brought by large models to the industry:
Searching for Everything in Literature: With just one word or sentence, you can instantly search billions of data points. For example, by entering "black backpack" or "red shirt", the system can accurately retrieve and locate the corresponding content, making it easy to search for "frames" in the ocean and no longer having to flip through frames to find the target image;
AI sports coach machine: "Smart Coach" captures human body posture in real time, accurately analyzes motion norms, and provides voice guidance, making sports fun and scientific. While reducing the burden of manual evaluation on teachers, it effectively improves teaching efficiency and quality;
Gas' Magic Mirror ': Through TDLAS technology and gas cloud imaging technology, viewers can' see 'harmful gas leaks that are invisible to the naked eye, and experience how AI safeguards industrial safety
Hikvision booth number: H2-E111, Shanghai World Expo Exhibition&Convention Center. We look forward to meeting you in Shanghai and embarking on an intelligent journey together.
