Trafic Sign Detection and Navigation System for Visually Impaired Individuals

In the context of modern transportation, accurate traffic sign prediction is quite crucial to guarantee efficient traffic management and safeguard of property and life. This work offers a new traffic sign and navigation system for artificial intelligence users with visual disabilities, so enhancing traffic safety and navigation accessibility. The basis of the system is RetirevaNet-43, a specifically built neural network meant to split traffic signs into $\mathbf{4 3}$ different groups. Unlike conventional object detection systems such the Yolo V8/V9/V11 optimized for general detection and localization, the RetirevaNet-43 is especially tailored to traffic sign recognition, so resulting in greater classification accuracy and faster inference times in this special domain. Combining two separate folding blocks, the RetirevaNet-43 employs a modular design. This design minimizes computing efforts in addition to guaranteeing correct identification of traffic signs. This implies that the system is fit for actual use in surroundings connected to resources. Modern intertext modules (TTS) also help to clear, natural monitoring feedback from classification. Users who are visually impaired should find this function rather valuable. This provides crucial data traffic signals and instantaneous speaking directions, so enhancing navigation. The YoroV8/V9/V11 of the system is exceeded in conditions requiring fine-tuned traffic sign classification and fast response times, according to a thorough experimental review. This paper emphasizes the opportunities of Taker AI systems to solve some real problems and open the path for easily available and more intelligent navigational artificial intelligence.