Smart Vision for the Visually Impaired


Presently, smart vision solutions are available for visually impaired persons, however, they are quite expensive and not commercially nor economically viable. This project aims to develop an intelligent object detection system which will categorize various objects that one tends to come across in daily living. Most of the currently available support tools for vision impairment are based on obstacle avoidance rather than object detection and categorization. This project aims to design a dedicated deep neural architecture, which will be employed in the form of a portable, wearable device consisting of tiny video cameras. The user will be able to know the type and distance of an object around them and be able to live more smartly with the help of this visual aid.

Where current solutions are available for anywhere between $1500 to $10,000, this device is intended to be priced around $100. The industrial partners for this project include Acrux Technologies and Pakistan Association of the Blind (PAB).

ThemeIoT, AI / Big DataTypeEdu-Tech / Start-up /