Enhancing Road Safety and ITS Efficiency: A Comprehensive Evaluation of Traffic Sign Detection using YOLOv8

Smart transportation systems (ITS) and autonomous vehicles hereby require up-to-date, highly accurate, and essential tools for sign detection with efficiency. This research paper presents a detailed evaluation of the You Only Look Once version 8 (YOLOv8) deep learning model for the detection and classification of traffic signs into four pivotal categories: red lights, stop signs, speed cameras, as well as crosswalks. The current study introduces a dataset from the Kaggle repository for a thorough investigation of model performance that simulates practical situations in which the weather is unstable and advanced signs vary in shape. The YOLOv8 model used sophisticated methodologies, which encompassed dataset preparation, model training, and solid evaluation metrics; which at the end of the day, showed that this model was better in sensitivity, specificity, and f1-scores, and at the same time, enabled real-time processing that applies to mobility and traffic management. It shows that YOLOv8 is much better than all the previous standard algorithms, as well, it is placed to be one of their best solutions in traffic sign detection. Regularly, conducted by the researchers contribute to the academic field because they present real-world evidence of the YOLOv8 model influence as well as providing practical implications for the future development of road safety and the operational efficiency of ITS and autonomous vehicles. The occurrence of such a location sets the scene for the development of novel traffic sign detection techniques, stressing the prospects of inputting improved approaches and the compilation of heterogeneous data to help in the categorization of the complex surrounding driving environments.