This project focuses on developing a real-time road sign detection system using deep learning and computer vision technologies. The system is designed to enhance the mobility and independence of visually impaired individuals by accurately identifying and interpreting road signs and providing real time audio feedback.
In this project, we utilized the TensorFlow 2 framework with CNN EfficientNet and leveraging SSD EfficientDet-D0 for training and object detection. For text-based signs, we integrated Keras-OCR to accurately convert the sign's text into a readable format. This combination of tools allows for precise and efficient road sign detection, enhancing the system's overall effectiveness for visually impaired users.