Enhanced Traffic Sign Recognition with Ensemble Learning

With the growing trend in autonomous vehicles, accurate recognition of traffic signs has become crucial. This research focuses on use convolutional neural networks for sign classification, specifically utilizing pre-trained models ResNet50, DenseNet121, and VGG16. To enhance accuracy robustness model, authors implement an ensemble learning technique with majority voting, to combine predictions multiple CNNs. The proposed approach was evaluated three different datasets: German Traffic Sign Recognition Benchmark (GTSRB), Belgium Dataset (BTSD), Chinese Database (TSRD). results demonstrate efficacy approach, rates 98.84% GTSRB dataset, 98.33% BTSD 94.55% TSRD dataset.