Traffic sign detection and recognition using Convolutional Neural Networks

This paper investigates a sophisticated framework based on convolutional neural networks (CNNs) for the real-time detection and recognition of traffic signs, utilizing the German Traffic Sign Recognition Benchmark (GTSRB) dataset. Traffic signs play a vital role in providing critical information for both human drivers and autonomous driving systems, ensuring the safe navigation of roads. A robust and accurate detection system is necessary for any autonomous vehicle to make decisions in real time. This work proposes an optimized CNN model that not only improves the recognition accuracy but also enhances the processing speed required for real-time applications. The proposed model builds on state-of-the-art architectures such as VGG-16 and Faster R-CNN to enable accurate traffic sign classification and real-time detection, demonstrating superior performance on the GTSRB dataset. By incorporating additional data augmentation techniques, batch normalization, and dropout, the model achieves a detection accuracy of 98.7% and a mean average precision (mAP) of 89.3%, marking a significant improvement over existing methods. Keywords: Traffic Sign Detection, Convolutional Neural Networks, GTSRB, Deep Learning, Image Recognition