An Assist System for Visually Impaired at Indoor Residential Environment using Faster-RCNN

Abstract

For the visually impaired, "independent living" is the key to rebuild the dignity and self-confidence. In addition to mobility skill, visually impaired memorize where the furniture is in homes. Therefore, moving the furniture or items (e.g. keys, remote control, etc.) around could be confusing and possibly unsafe. This article describes a system that combines deep learning with a camera for visually impaired to identify objects in the space whether it has been moved, such as furniture, important daily necessities, etc. The smart phone app is used to inform the visually impaired about the current location and movement location of all objects. We hope to re-establish a safe, convenient and comfortable indoor environment for the visually impaired through our system.