Aura Assist
Loading...
Date
Journal Title
Journal ISSN
Volume Title
Publisher
St. Francis Institute of Technology (SFIT)
DOI
Abstract
Aura Assist is an assistive system specially designed to support visually impaired individuals in moving around more safely and independently. Traditional aids like white canes or guide dogs are helpful but limited, they can detect obstacles in close range but cannot describe the environment or recognize people. Aura Assist fills this gap by using modern computer vision and deep learning techniques to turn visual information into meaningful audio feedback. The system uses a webcam to continuously scan the environment and provides voice alerts about objects, people, and navigation paths. This helps users build a mental map of their surroundings and reduces the risks associated with limited vision. One of the core strengths of Aura Assist is its ability to detect objects and recognize faces in real-time. It uses YOLOv8, a fast and accurate deep learning model, to detect obstacles like furniture, walls, doors, or other items that could block the users path. The system immediately announces what it sees using a text-to-speech engine, allowing users to avoid collisions and better understand their environment. For face recognition, it employs MTCNN for face detection and InceptionResNetV1 to identify known individuals. When a registered face is detected, the systemsays the persons name aloud. This feature is especially helpful in social situations, making it easier for visually impaired individuals to interact with others confidently. Over time, users can add more familiar faces, making the experience more personal and effective. In addition to object and face recognition, Aura Assist includes a voice-controlled indoor navigation system. Users can simply speak where they want to go, and the system uses speech recognition and a graph-based shortest path algorithm to calculate the best route. It then provides step-by-step voice instructions to help the user move safely through hallways, rooms, or buildings. The whole system is developed as a Flask-based web application, ensuring it runs smoothly on various devices and platforms. Its interface is clean, responsive, and easy to use. Aura Assist brings together advanced AI technologies into one smart system that not only improves mobility and spatial awareness but also strengthens social interaction and independence. By empowering users with real-time information about their surroundings, Aura Assist aims to significantly improve their quality of life.
Description
Project Report submitted to St. Francis Institute of Technology (SFIT).
Citation
Assistive Technology
Visually Impaired
Object Detection
Face Recognition
YOLOv8
MTCNN
InceptionResNetV1
Computer Vision
Deep Learning
Text-to-Speech
Indoor Navigation
Visually Impaired
Object Detection
Face Recognition
YOLOv8
MTCNN
InceptionResNetV1
Computer Vision
Deep Learning
Text-to-Speech
Indoor Navigation