Paper Title
DYNAMIC TRAFFIC LIGHT SYSTEM USING YOLOV3 AND KNN REGRESSION

Abstract
Urban traffic management has become increasingly complex, necessitating the development of advanced methods to optimise traffic flow and mitigate congestion. Traditional systems with fixed signal timings fail to address dynamic traffic scenarios effectively. In this paper, we propose a system integrating YOLOv3 for vehicle detection and K-Nearest Neighbors (KNN) regression to dynamically allocate traffic signal times based on real-time vehicle counts. Our approach offers real-time adaptability, computational efficiency, and scalability, providing a practical solution to enhance urban traffic management.