3D Scene Description Using Multimodal Models for Assisting Blind People in Outdoor Navigation

Institut

Professur für autonome Fahrzeugsysteme

Typ

Masterarbeit /

Inhalt

experimentell / theoretisch / konstruktiv /  

Beschreibung

Also possible as IDP, research internship (Forschungspraxis), guided research, etc.

This research focuses on leveraging multimodal models to create detailed 3D scene descriptions to assist blind and visually impaired individuals in outdoor navigation. The project aims to integrate data such as color and depth images, as well as outputs from AI models (e.g., segmentation), to generate accurate 3D representations of the environment. These representations will provide context-aware guidance and 3D scene descriptions tailored to the needs of blind users.

The project involves researching and evaluating existing methods, optimizing them, and tailoring them to the specific needs of blind individuals. Additionally, the system's effectiveness will be assessed through field testing with visually impaired users to ensure real-world applicability and usability.

This research is conducted in collaboration with the Startup SmartAIs
https://de.linkedin.com/company/smartais

Voraussetzungen

• Strong knowledge of C++ and Python
• Experience with 3D computer vision and AI-based segmentation methods
• Familiarity with machine learning and multimodal data integration
• Independent and organized work attitude
• Passionate about accessibility and assistive technologies

 

Möglicher Beginn

sofort

Kontakt

Yamen Mohisn
yamen.mohisnsmartais.de