2D Image to 3D Mesh Experiments

Since 2022, I have been actively engaged in exploring the generation of AI-based architectural renderings from textual prompts. My work has focused on understanding and advancing how text-to-image AI models can support early-stage architectural visualization. Over this period, I have closely followed the development of the AI 3D projects, which has now reached a level of maturity where it is possible to generate highly complex 3D models directly from AI outputs. These models can be exported into professional design environments such as Blender and Rhinoceros 3D, where they can be scaled, measured, and refined. This capability allows the AI-generated geometry to serve not only as a visual concept but also as a practical starting point for architectural design workflows. By bridging the gap between AI-rendered imagery and usable 3D geometry, this work has the potential to accelerate iterative design processes and introduce new ways of integrating generative AI into architectural practice.