Dahua Technology presenta los modelos de IA a gran escala Xinghan
Water Technologies

Dahua Technology presenta los modelos de IA a gran escala Xinghan

-S Constant Volution for the future of Aiot Intelligent: Dahua Technology presents the Xinghan large -scale AI models

HANGZHOU, China, September 21, 2025 / PRNEWSWIRE/ – DAHUA TECHNOLOGY, world leader of IAOT video -centered solutions and services, has officially launched its Xinghan large -scale AI modelsa latest generation and industrial level system that integrates large -scale visual intelligence with multimodal and linguistic capacities. Developed to address the complex challenges of real environments, Xinghan represents a great advance in the continuous innovation of Dahua, promoting intelligent transformation in various sectors.




The Dahua Xinghan Large-Scale AI Model places visual analysis at its core, integrating multimodal capabilities and embedding deep industry expertise to create a large model tailored for diverse application scenarios. Built upon this real-world foundation, it serves as the key to achieving scalable and commercially viable AI solutions.






The Dahua Xinghan Large-Scale AI Model places visual analysis at its core, integrating multimodal capabilities and embedding deep industry expertise to create a large model tailored for diverse application scenarios. Built upon this real-world foundation, it serves as the key to achieving scalable and commercially viable AI solutions.

Xinghan Technological Foundation

With the mission of allowing machines to truly understand the world, the Xinghan models system continues to evolve, connecting avant -garde research with real applications. Baptized with the name of the Chinese word for “galaxy”, Xinghan offers an integral capacity matrix driven by the synergy between the edge and the cloud, which allows a scalable and adaptive intelligence in all sectors. Xinghan’s improved architecture consists of three series of main models: L, V and M. The L -series model focuses on the understanding and interaction of natural language, while the other two address more specific applications:

V series: Xinghan vision models

Focused on advanced visual intelligence and video analysis, this series optimizes the target categories focusing on key objectives (for example, humans, motorized and non -motorized vehicles) to reduce the complexity of the model and maintain high precision.

Main characteristics:

  • Perimeter protection: Coverage is extended by accurately identifying smaller objectives (even up to 20 × 20 pixels) compared to CNN -based traditional AI models, which reduces false alarms and increases the detection range of large models cameras.*
  • WizTracking: It offers a state -of -the -art smart monitoring algorithm that can manage complex occlusion and variations in the position of the objective, achieving a 50 % improvement in precision.*
  • Map of multitudes: It significantly improves the detection of small objectives at long distances (up to twice from distance) and includes umbrella compensation, which improves accuracy at 80 % in rain conditions*. In addition, it offers an increase of 2.5 times in the analysis range, admits the detection of up to 5,000 people and offers robust yield in dense crowds and environments with little light.*
  • Adaptable to the scene – WDR con IA: Take advantage of the knowledge of the situation to analyze the spatial and contextual characteristics of a scene, allowing an intelligent and automated configuration of the camera.
  • Rules assistant with AI: Designed for automatic delineation of perimeter protection intrusion rules, it offers access with a single click, recognition of high precision scenes, automatic analysis and much more.

M series: Xinghan multimodal models

Multimodal models are advanced systems capable of processing simultaneously and integrating multiple types of heterogeneous data (eg, text, images, audio and video). This significantly improves the efficiency of information processing, allows a more natural person-order interaction and opens a broader spectrum of application scenarios.

Main characteristics:

  • WizSeek: Revolutionary video research through natural language search. Simply describe your objective (for example, people, vehicles, animals or objects) and Wizseek instantly recovers the coincident recordings in video files.
  • Alarms defined by text: It allows users to define alarms simply describing them in natural language, which significantly reduces the development threshold and allows rapid, flexible and scalable configuration adapted to various real world scenarios.

For more information about Xinghan’s large -scale AI models, visit here.

* The results are based on a standard configuration and test environment ..

Photo – https://mma.prnewswire.com/media/2777186/image.jpg
Logo – https://mma.prnewswire.com/media/2074657/Dahua_LOGO_Logo.jpg