X-Part: high fidelity and structure coherent shape decomposition

TL;DR

The paper tackles the challenge of generating and decomposing 3D shapes into meaningful parts with high fidelity and controllability. It introduces X-Part, a diffusion-based framework that uses bounding-box prompts and P3-SAM semantic features to drive synchronized multi-part generation and provides an editable pipeline for interactive modifications. Through a vecset-based latent diffusion backbone and a Diffusion Transformer, it achieves state-of-the-art performance in part-level decomposition and holistic shape generation, with robust editing capabilities. The work promises production-ready, editable 3D assets and highlights practical benefits for downstream tasks like UV unwrapping and mesh retopology.

Abstract

Generating 3D shapes at part level is pivotal for downstream applications such as mesh retopology, UV mapping, and 3D printing. However, existing part-based generation methods often lack sufficient controllability and suffer from poor semantically meaningful decomposition. To this end, we introduce X-Part, a controllable generative model designed to decompose a holistic 3D object into semantically meaningful and structurally coherent parts with high geometric fidelity. X-Part exploits the bounding box as prompts for the part generation and injects point-wise semantic features for meaningful decomposition. Furthermore, we design an editable pipeline for interactive part generation. Extensive experimental results show that X-Part achieves state-of-the-art performance in part-level shape generation. This work establishes a new paradigm for creating production-ready, editable, and structurally sound 3D assets. Codes will be released for public research.

Figures (6)

• Figure 1: Architecture of $\mathcal{X}$-Part. Given input point cloud, per-point feature and part bounding boxes are extracted from $\text{P}^3\text{-SAM}$. Global and part conditions are obtained by stacking geometry token with interpolated semantic features. They are injected to multi-part diffusion process to guide shape decomposition.
• Figure 2: Qualitative shape decomposition results. Note the input shapes for HoloPart and Ours are obtained from Hunyuan3D-2.5 lai2025hunyuan3d, while OmniPart leverage the shape from trellis xiang2025structured.
• Figure 3: Qualitative shape decomposition results. Note the input shapes for HoloPart and Ours are obtained from Hunyuan3D-2.5, OmniPart leverage the shape from Trellis, PartCrafter and PartPacker are image to 3D methods, they do not rely on shapes.
• Figure 4: Demonstration of two representative applications of our method. Subfigure (a) shows the results of bounding box-controlled part generation, while subfigure (b) illustrates improved UV unwrapping performance achieved through part-based decomposition.
• Figure 5: Part generation results under different module ablation settings.