Journal of Computational Design & Engineering
accepted version
Title A multi-modal workflow for image-to-3D generation and multi-objective optimisation of generated 3D models Authors (1,2) Kapsalis, E.; (1) Jaeger, N. (1): University of Nottingham, UK (2): University of Derby, UK
0. Abstract Form-finding and optimisation tools enable architects and engineers to draw inspiration from precedent projects and generate a multitude of design alternatives, based on certain criteria, during the early stages of design. Despite the recent advancements in the fields of generative AI and multi-objective evolutionary computation, there remains significant potential for further development and integration of these techniques in the design exploration process. In this paper, we propose a new technique of generating 3D models from 2D images or technical drawings, and subsequently optimising their design based on sets of multiple objectives. We also present a case study to demonstrate the practical applicability and effectiveness of the proposed workflow. Results from the case study indicate the usability of our approach in generating a diverse set of design alternatives, optimising them based on multiple criteria, and ultimately selecting a design solution that is well-aligned with project goals and constraints. We conclude by discussing the practical, technical, and ethical implications of our approach. Keywords: generative AI, image-to-3D, evolutionary algorithms, multi-objective optimisation, design space exploration, form-finding