EchoVision: RealTime Object Detection and Voice Assistance for the Visually Impaired

Visual impairment is one of the greatest challenges that individuals must overcome to access and navigate their environment. To address this crucial issue, the paper discusses an intelligent system that aims to improve the accessibility and independence of visually impaired people in their local communities. The proposed system uses an object detection model that interfaces with a text-to-speech system. It enables the user to express orally the names of objects which have been detected in their selected language and provides relevant information in the user’s native language. Through a comprehensive evaluation, the efficiency of this system has been carefully examined. The results show important improvements on the object detection score which will enable users to interact with their environments with greater confidence and more independence. This paper takes an important step in narrowing the access gap and bringing independence to visually impaired individuals.