RAG-Based Model Generation for 3D Printing Through Code Synthesis
摘要
In recent years, Additive Manufacturing (AM;3D Printing), has experienced considerable growth as an alternative to traditional Subtractive Manufacturing (SM) methods, particularly for prototyping, but increasingly in largescale commercial applications. Advances in printing technology and reduction in material costs have driven this shift. Despite these advancements, challenges associated with generating printable three-dimensional models remain. This process can often be time-consuming and necessitates a high level of expertise. Consequently, there has been a marked interest in integrating generative artificial intelligence (AI) technologies to automate 3D model generation. This paper aims to develop a parameter-aware generative AI framework that synthesizes hollow 3D models of boxes, cylinders, cones, spheres and pyramids from textual descriptions while incorporating specifications for printability.