Quartet of Diffusions: Structure-Aware Point Cloud Generation through Part and Symmetry Guidance

TL;DR

The paper presents Quartet of Diffusions, a novel structure-aware point cloud generation framework that models shape parts and symmetry explicitly, enabling high-quality, controllable 3D shape synthesis with guaranteed symmetry and coherence.

Contribution

It introduces a four-model diffusion pipeline that explicitly encodes shape parts, symmetry, and global structure, a first in fully integrating these priors in 3D point cloud generation.

Findings

Achieves state-of-the-art performance in point cloud generation.

Ensures guaranteed symmetry and coherent part placement.

Supports fine-grained control over shape attributes.

Abstract

We introduce the Quartet of Diffusions, a structure-aware point cloud generation framework that explicitly models part composition and symmetry. Unlike prior methods that treat shape generation as a holistic process or only support part composition, our approach leverages four coordinated diffusion models to learn distributions of global shape latents, symmetries, semantic parts, and their spatial assembly. This structured pipeline ensures guaranteed symmetry, coherent part placement, and diverse, high-quality outputs. By disentangling the generative process into interpretable components, our method supports fine-grained control over shape attributes, enabling targeted manipulation of individual parts while preserving global consistency. A central global latent further reinforces structural coherence across assembled parts. Our experiments show that the Quartet achieves state-of-the-art performance. To our best knowledge, this is the first 3D point cloud generation framework that fully integrates and enforces both symmetry and part priors throughout the generative process.