Third Eye: A Smart Assistant for Visually Impaired using Deep Learning
T. R. Chenthil · International Journal for Research in Applied Science and Engineering Technology · 2021
Globally, one billion people have a vision impairment resulting in them unable to live a normal life.This project aims to develop a technological solution effectively enabling the visually impaired to identify text and objects in front of them.This Third Eye based on Raspberry Pi uses Deep Learning (DL) which tends to support in identifying the objects and faces that helps in orientation and navigation of these people in the unknown surrounding.It can also recognize text or any readable data and recite using earphones or speakers.The system will perform voice recognition from the user, when the user states object mode the camera is initiated enabling live streaming and a bounding box is applied to the object from which the distance and direction are calculated, then a voice note is given stating the name of the object, distance, and direction.If a face is detected, then it recognizes the face using the Detection Transformer (DETR) algorithm and then says the name of the person detected.When the user states text mode, the camera is initiated enabling live streaming and the text is captured and a voice is given stating the text.Thus, providing an effective technological solution for the blind to effectively identify objects and text in front of them.The proposed system which is fully voice-controlled uses the Raspberry Pi as a controller.The smart assistant is user-friendly and has a quick response which ensures real-time object detection for the visually impaired.