Vision Beyond Sight: An AI-Assisted Navigation System in Indoor Environments for the Visually Impaired

This study introduces an innovative visual assistance technology that incorporates a cutting-edge live object recognition system to address the prevalent challenges faced by visually impaired individuals worldwide. In daily life, the visually impaired rely heavily on touch and auditory cues to navigate their surroundings, which underscores the urgent need for technological support. While existing solutions such as Screen Reading software and Braille devices fail to recognize crucial elements, rendering them ineffective in critical situations, our approach aims to revolutionize the lives of the blind and visually challenged on a global scale by providing a second set of eyes without external assistance. With a commitment to creating an inclusive environment, our research goes beyond current solutions by emphasizing assistive technology, which truly makes a difference. The proposed system leverages SSD MobileNet-v2 FPNlite-based object detection combined with an Android mobile application to empower visually impaired individuals through real-time voice-based guidance. By specifically addressing the challenges related to recognition and navigation, this project aims to significantly improve the independence and overall quality of life of those with visual disabilities, thereby contributing to the creation of a more welcoming and inclusive world.