A Comparative Deep Learning Approach for Vehicle Speed Monitoring using YoLov11

This study investigates the adoption of deep learning techniques in order to achieve the intelligent monitoring as well as analysis of vehicle speeds. The primary target is to look into vehicle detection using some of the deep learning model belonging to the YoLo series. This study considers therefore the use of YoloV11, the latest model in the YoLo series, and the description below outlines how these architectural improvements have improved detection efficiency and robustness with greater speed especially in complicated and blocked environments. The model is then compared with other deep learning techniques like Convolutional Neural Network (CNN), Long Short-Term Memory (LSTM), EfficientNetB0 using AI City Challenge dataset, which is a challenging set of vehicle speed reading images. The data contains a total of 4,168 annotated images from different kinds of vehicles, which represent an example of the various types of vehicles found on traffic systems. The results indicated that the YOLOv11 model has an accuracy of 98.8% and an F1 score of 99.2% more than the CNN model, which has an accuracy of 93.3% and has an F1 score of 95.1%. The LSTM model and the EfficientNetB0 give accuracy scores of 95.7% and 96.5% respectively. The focus of this study is the application of some of the deep learning methods in vehicle speed prediction and anomaly detection, and these strengthen the main issue of intelligent transportation systems.