Safe-Sight

Authors: Parth Bhalekar, Aarya Darne, Sakshi Devadiga, Mrs. Akanksha Patil

Abstract: The absence of immediate visual data presents significant obstacles for individuals with visual impairments when navigating unfamiliar settings. Although they provide some degree of support, conventional assistive technologies, such as white canes, guide dogs, and basic smartphone applications, often exhibit limitations in terms of functionality, cost, or real- time adaptability. To address these limitations, this study proposes Safe-sight, a mobile application using artificial intelligence. The goal is to improve the safety and independence of people with visual impairments. The system uses text-to- speech technology to provide immediate audio feedback. It also uses optical character recognition to read text, and object detection, which is powered by deep learning models. Voice command is the simplest way to interact, allowing for use and easy control. The result highlight the system's ability to improve mobility, situational awareness, and confidence for users with visual impairments. Safe-sight, therefore, provides a solution that's both practical and scalable, while also being budget- friendly, merging accessibility with the latest AI developments.