Abstract

This study delves into the comparative performance analysis of three consecutive iterations of the You Only Look Once (YOLO) object detection framework—YOLOv8, and YOLOv9—specifically tailored for the challenges posed by a dedicated traffic signs dataset. The objective is to offer a thorough comprehension of the advantages and disadvantages that each version concerning traffic sign detection, with implications for advancing real-world object detection in dynamic traffic scenarios. Preliminary findings underscore YOLOv9's standout performance, demonstrating superior precision and F1-score when contrasted with YOLOv7 and YOLOv8. This observation accentuates YOLOv8's potential for delivering robust traffic sign detection solutions. The implications of our research extend beyond benchmarking, contributing essential knowledge to the field of computer vision applied to traffic sign recognition. The study's outcomes serve as a valuable resource for practitioners and researchers seeking to select an optimal model for similar applications, ultimately fostering advancements in intelligent transportation systems.