I. Introduction
This article delves into the challenge of designing an algorithm for recognizing individuals drowning in coastal areas. Drowning occurs when an individual's air supply is obstructed, preventing normal breathing. Such incidents typically transpire in pools or at beaches. While most public pools and beaches are staffed with professional lifeguards, drowning or near-drowning events still occur due to the reaction time of the responders. This underscores the pressing need to automate the detection of such critical incidents. One promising approach to this automation is the deployment of surveillance cameras equipped with computer vision algorithms tailored to identify drowning instances in pools and beachfronts [1], [2].