Skip to content
Neural Concept Connect 2026 · Bay Area ✓ · Next: Munich (Nov 24) · Tokyo (Feb 3, 2027) Apply for your spot
Neural Concept
Get started

HybridSDF: Combining Free Form Shapes and Geometric Primitives for effective Shape Manipulation

In short

CAD modelling relies on simple geometric primitives, while deep-learning 3D surface modelling opened new design possibilities that the CAD community has not adopted because they do not fit engineering workflows. HybridSDF combines primitives with free-form implicit surfaces, keeping accuracy, interpretability and consistency while remaining easy to manipulate.

Key takeaways

  • The blocker for learned 3D surface modelling in CAD is workflow integration, not raw modelling capability.
  • Combining geometric primitives with implicit free-form surfaces is the proposed way to bridge the two.
  • Interpretability and consistency are treated as requirements, because engineers must be able to reason about the shape.
  • Easy manipulation is the practical test, since a representation engineers cannot edit will not be adopted.

CAD modeling typically involves the use of simple geometric primitives whereas recent advances in deep-learning based 3D surface modeling have opened new shape design avenues. Unfortunately, these advances have not yet been accepted by the CAD community because they cannot be integrated into engineering workflows. To remedy this, we propose a novel approach to effectively combining geometric primitives and free-form surfaces represented by implicit surfaces for accurate modeling that preserves interpretability, enforces consistency, and enables easy manipulation.

Read the full article here.