Light-SQ: Structure-aware Shape Abstraction with Superquadrics for Generated Meshes

2025-09-29

AI Summary

To address the challenge of editing high-resolution generative meshes by non-professional users in UGC scenarios, this paper proposes a structure-aware superquadric shape abstraction framework that compresses meshes into compact, editable primitive representations. Methodologically: (i) Signed Distance Function (SDF)-guided carving suppresses primitive overlap; (ii) a voxel-based block-wise regrowth strategy enables structured part segmentation; (iii) adaptive residual pruning prevents over-segmentation; and (iv) multi-scale superquadric optimization preserves fine geometric details. Evaluated on the newly constructed 3DGen-Prim benchmark, our method significantly improves structural plausibility (low overlap, part alignment), geometric fidelity, and editability of abstractions. It constitutes the first structure-aware abstraction solution explicitly designed for user-driven editability in generative 3D content creation.

Technology Category

• Computer Vision: 3D Computer Vision
• Machine Learning: Deep Generative Models & Autoencoders
• Humans and AI: Game Design — Procedural Content Generation & Storytelling

Application Category

• Web Mining and Content Analysis: Bridging structured and unstructured data
• Economics, Online Markets and Human Computation: Economic ramifications for generative AI infrastructure and applications
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs

Abstract

In user-generated-content (UGC) applications, non-expert users often rely on image-to-3D generative models to create 3D assets. In this context, primitive-based shape abstraction offers a promising solution for UGC scenarios by compressing high-resolution meshes into compact, editable representations. Towards this end, effective shape abstraction must therefore be structure-aware, characterized by low overlap between primitives, part-aware alignment, and primitive compactness. We present Light-SQ, a novel superquadric-based optimization framework that explicitly emphasizes structure-awareness from three aspects. (a) We introduce SDF carving to iteratively udpate the target signed distance field, discouraging overlap between primitives. (b) We propose a block-regrow-fill strategy guided by structure-aware volumetric decomposition, enabling structural partitioning to drive primitive placement. (c) We implement adaptive residual pruning based on SDF update history to surpress over-segmentation and ensure compact results. In addition, Light-SQ supports multiscale fitting, enabling localized refinement to preserve fine geometric details. To evaluate our method, we introduce 3DGen-Prim, a benchmark extending 3DGen-Bench with new metrics for both reconstruction quality and primitive-level editability. Extensive experiments demonstrate that Light-SQ enables efficient, high-fidelity, and editable shape abstraction with superquadrics for complex generated geometry, advancing the feasibility of 3D UGC creation.

Problem

• Compressing high-resolution meshes into compact editable representations
• Achieving structure-aware shape abstraction with minimal primitive overlap
• Enhancing primitive-level editability for 3D user-generated content creation

Innovation

• SDF carving reduces primitive overlap iteratively
• Block-regrow-fill strategy guides structural partitioning
• Adaptive residual pruning suppresses over-segmentation compactly