AI model lightweighting and optimisation company Nota said on Sunday it has upgraded its video monitoring solution, Nota Vision Agent (NVA), based on Nvidia’s Video Search and Summarization (VSS) framework.

NVA is a solution that understands situations in video based on a vision language model and supports operators so they can search, summarise and report needed information in natural language. Nota, an Nvidia certified partner, provides customer-tailored functions by incorporating VSS into NVA.

Nota has deployed NVA in the traffic control system of the Daejeon Regional Construction and Management Administration. It detects unexpected situations such as accidents, fires and obstacles in real time through road CCTV, and automatically summarises and reports lane-by-lane traffic information and response status.

In industrial safety, it applied NVA to Kolon Industries’ Gimcheon Plant 2 together with Kolon Benit. It is used for worker safety monitoring, surveillance of hazardous areas and screening for violations of safety rules. Nota said VSS-based video search and summarisation help quickly identify dangerous situations.

Nota CEO Myungsoo Chae (채명수) said the video monitoring market is rapidly shifting beyond simple object detection to generative AI-based video AI agents. He said Nota would expand verified on-site cases and, together with Nvidia, broaden the global video AI agent market.