
SenseFoundry Algorithm Marketplace
-

With“Visual AI + Multimodal Large Models"
driving Smart City and Commercial Intelligence TransformationAfter a decade of intensive reseach and developement, SenseFoundry leverages AI algorithms to deliver a powerful visual AI analysis and processing engine, focusing intensely on smart cities and commercial domains. Now driven by the innovative integration of “visual AI + large models + intelligent agents,” it can comprehend and process complex information across multiple scenarios, efficiently meeting the visual algorithmic task demands of various businesses in vertical markets. Currently, SenseFoundry has been deployed in nearly 200 cities across the Asia-Pacific region, including 16 megacities with populations exceeding 10 million and 4 overseas cities, covering nearly 30,000 campuses, buildings, retail outlets, and transportation hubs.
-

In terms of technical support, SenseFoundry (SenseTime Ark) provides robust underlying capabilities and is positioned as an open visual empowerment platform that can scale to support the integrated processing and analysis of 100,000+ video feed sources and 100-billion-level unstructured features and structured information. Its sixth-generation version, "Ark VL," incorporates multi-modal foundation models, deeply integrating multi-dimensional capabilities such as urban basic algorithms, urban scenario-specific algorithms, multi-modal foundation model algorithms, and AI agents. This integration effectively addresses key pain points faced by traditional urban visual services, including oversimplified interaction methods, inability to interpret generalized scenarios, high false alarm rates in complex scenarios, and difficulty in adopting foundation model applications.
At the technological core lies the SenseFoundry open platform, a scalable visual intelligence infrastructure capable of processing up to 100,000 video streams and integrating hundreds of billions of unstructured and structured data. The sixth-generation "Ark VL" version incorporates multimodal large models, merging foundational algorithms, scenario-specific models, AI agents, and more. This addresses key challenges in traditional urban vision systems—such as limited interaction modes, poor generalization in complex scenarios, high false alarm rates, and steep learning curves for large model adoption. The Sense Foundary serves diverse sectors including public service intelligence, emergency response, urban decision support, regional development, and intelligent transportation. It also empowers commercial intelligence transformation. By leveraging SenseTime's self-developed visual large models, the platform provides not only general algorithms, scenario-specific solutions, OCR recognition, and fusion retrieval for surveillance but also advanced multimodal capabilities. These include multimodal event detection, retrieval, open-event identification, urban intelligent tagging, and secondary filtering—expanding technical boundaries and delivering smarter, more efficient services.
Why Choose Us
Milestones

Our Vision
To maintain technological leadership in the field of visual AI
Contact Us
Contact Us
-
Phone Inquiry
400 900 5986(Monday to Friday, 9:00-12:00, 13:00-18:00) -
Partnership Email:
sensefoundry_service@sensetime.comMedia Inquiries:
pr@sensetime.com
Follow us on WeChat Official Account

