Video Analysis of Drug & Human Smuggling Activities

During this Phase 1 SBIR project for the Department of Homeland Security (DHS), Vadum developed Alert Handling for Anomalous Behavior (AHAB), a pioneering machine learning (ML) system for automatic analysis of maritime video surveillance feeds to detect and flag multifaceted behaviors associated with drug smuggling. 

dhs uscg sbir

AHAB utilizes a novel multi-scale spatial and temporal analytics engine to detect salient, complex, and evolving behaviors of interest. By providing timely alerts, valuable insights, and alerting rationale to the United States Coast Guard (USCG) watchstander. AHAB helps alleviate watchstander cognitive burden, reduce fatigue, and improve decision making to increase successful interdictions. To accommodate the ever-changing tactics of drug and human smugglers, the system is designed to receive and incorporate operator feedback, enabling fine-tuning of alerts and expansions to features of interest. In addition to Coast Guard applications, AHAB is easily adaptable to domains such as border patrol surveillance or monitoring of military outposts, government facilities, and critical public infrastructure.