Sight Seeing with Wearable Tech: A Ray of Hope for the Visually Impaired

Object Detection & Categorization for Blind using Deep Neural Learning

Summary

OPTImization & MAchine Learning (OPTIMAL) lab is developing wearable glasses to assist the visually impaired people in interacting with the environment using deep neural networks.

This project aims to develop a portable/wearable device for the visually impaired to detect and recognize general objects close to the user in a video stream using stereo cameras and miniature computers.
The solution will improve upon image processing techniques for depth analysis in a given video stream. This will be done using a deep neural architecture for scene segmentation and finally object recognition in the image or video frame. Furthermore, it will also port deep neural networks and various image processing approaches to a portable computer like Raspberry Pi or the Odroid with a stereo camera interface.
The system will communicate with the user using speech signals for detected / recognized objects in the video that are at a certain distance from the user.

ThemeWearables/Implantable, AI/Big DataTypeHealth-Tech/AcademiaWebsiteoptimal.seecs