Study of EfficientNet for Traffic Sign Recognition

Lim, Xie Han (2025) Study of EfficientNet for Traffic Sign Recognition. Final Year Project (Bachelor), Tunku Abdul Rahman University of Management and Technology.

Text LimXieHan - FULL TEXT.pdf Restricted to Registered users only Download (3MB)

Abstract

This study investigates the application of the EfficientNet model to traffic sign recognition (German Traffic Sign Recognition Benchmark (GTSRB), China Traffic Sign Recognition (CTS) datasets), and the Malaysia Traffic Sign Recognition (MTS), and focuses on their accuracy and computational efficiency in autonomous vehicles. Traffic sign recognition is critical to ensuring the safety and efficiency of autonomous vehicles, which must accurately recognize and interpret a variety of road signs under different environmental conditions. Traditional deep learning models, such as convolutional neural networks (CNNs), require a lot of computational resources, which poses a challenge for real-time processing in resource-constrained environments such as autonomous vehicles. EfficientNet is a CNN architecture that addresses these challenges by optimizing network depth, width, and resolution to achieve state-of-the-art performance with fewer parameters and lower computational costs. The results show that the EfficientNet model can significantly improve the accuracy and efficiency of traffic sign recognition systems, contributing to the development of safer and more reliable autonomous vehicles.