Generative Multi-Modal Artificial Intelligence for Dynamic Real-Time Context-Aware Content Creation in Augmented Reality

We introduce a framework that uses generative Artificial Intelligence (AI) for dynamic and context-aware content creation in Augmented Reality (AR). By integrating Vision Language Models (VLMs), our system detects and understands the physical space around the user, recommending contextually relevant objects. These objects are transformed into 3D models using a text-to-3D generative AI techniques, allowing for real-time content inclusion within the AR space. This approach enhances user experience by enabling intuitive customization through spoken commands, while reducing costs and improving accessibility to advanced AR interactions. The framework’s vision and language capabilities support the generation of comprehensive and context-specific 3D objects.