Summary
Creating computer programs that design 3D objects from text is tricky because the program needs to understand how the object is built, not just what it looks like. The authors developed a new method that represents a design’s intended parts and how they fit together in a tree structure called a Constraint Intent Tree (CIT). This tree helps guide the creation of the design program and checks if the program matches the intended design process. By comparing expected and actual constraints, their system can find and fix mistakes, making designs more accurate especially for complex objects. Their approach moves beyond just matching shapes to focusing on the construction steps behind the designs.
Computer-Aided Design (CAD)parametric programslarge language models (LLMs)geometry constraintsConstraint Intent Tree (CIT)Boolean operationsdesign intentcode verificationprogram synthesis
Authors
Yali Du, Hui Sun, San-Zhuo Xi, Ming Li
Abstract
Natural-language Computer-Aided Design (CAD) code generation aims to turn design intent into executable and editable parametric programs. Large language models (LLMs) make this goal increasingly practical, but useful systems must preserve the construction process behind the rendered geometry. Existing benchmarks and methods mostly focus on how closely the generated CAD model matches the reference geometry, often using metrics such as Intersection over Union (IoU). Such metrics can miss errors in part decomposition, construction hierarchy, Boolean operations, sketch structure, and geometric relations. This gap calls for a representation that makes design intent explicit and lets a system check generated code against that intent. We propose CIT-CAD, a framework that infers a Constraint Intent Tree (CIT) from the input description to represent the intended entities, hierarchy, operations, and relations. The tree has two roles: it guides CAD code generation and defines expected constraints for verification. The framework extracts actual constraints from the generated program, compares them with the expected constraints, and uses mismatches to localize and repair design violations. Experiments show that the framework improves CAD generation performance, with larger gains on more complex multi-entity designs. By turning design intent into an explicit and checkable object, this work is the first attempt to move text-to-CAD generation beyond rendered-geometry matching toward construction-aware synthesis, verification, and repair.