Traffic Sign Detection Using SSD Mobilenet & Faster RCNN

Abstract

This paper presents a deep learning-based approach to traffic sign detection. The proposed approach utilizes state- of-the-art object detection models, such as SSD and RCNN, to detect traffic signs in real time. Tensorflow is used as a platform for training deep learning models and the models are being implemented on Google Colab and Kaggle cloud due to the GPU availability on the respective platforms. The models are evaluated on a traffic sign dataset from Kaggle, and it achieves high detection accuracy. Moreover, the proposed approach is robust to different lighting and weather conditions and is capable of detecting traffic signs from various distances and angles. The findings of this study show that by effectively detecting and recognizing traffic signs, deep learning-based techniques can potentially increase the safety and effectiveness of transportation networks.