Videos

This solution leverages multi-modal Generative AI to generate concise video summaries for in-store theft prevention. Using a Vision-Language Model (VLM), it integrates visual data from IP cameras with embeddings from a Large Language Model (LLM) to detect critical events like theft, empty shelves, and spills. The AI driven summary enhances situational awareness, reduces manual video review efforts, and enables swift responses. Stored in a database, summaries can be indexed and retrieved for further analysis, optimizing security, inventory management, and overall store operations.