UltraShape 1.0：通过可扩展几何优化实现高保真三维形状生成

UltraShape 1.0: High-Fidelity 3D Shape Generation via Scalable Geometric Refinement

December 24, 2025

作者: Tanghui Jia, Dongyu Yan, Dehao Hao, Yang Li, Kaiyi Zhang, Xianyi He, Lanjiong Li, Jinnan Chen, Lutao Jiang, Qishen Yin, Long Quan, Ying-Cong Chen, Li Yuan

cs.AI

摘要

在本报告中，我们推出UltraShape 1.0——一个可扩展的高保真三维几何生成扩散框架。该方案采用两阶段生成流程：首先生成粗略的全局结构，随后进行细化以产生细节丰富的高质量几何体。为支撑可靠的三维生成，我们开发了包含新型水密处理方法和高质量数据过滤的综合数据处理流程。该流程通过剔除低质量样本、填补孔洞及增稠薄壁结构，在保留细粒度几何细节的同时，显著提升了公开三维数据集的几何质量。为实现细粒度几何优化，我们在扩散过程中将空间定位与几何细节合成解耦：通过基于体素的固定空间位置细化，利用粗粒度几何体导出的体素查询提供经RoPE编码的显式位置锚点，使扩散模型能够聚焦于在缩小的结构化解空间内合成局部几何细节。我们的模型仅使用公开三维数据集进行训练，在有限训练资源下仍实现了卓越的几何质量。大量评估表明，UltraShape 1.0在数据处理质量和几何生成能力上均与现有开源方法具有竞争优势。所有代码与训练模型将全面开源以支持后续研究。

English

In this report, we introduce UltraShape 1.0, a scalable 3D diffusion framework for high-fidelity 3D geometry generation. The proposed approach adopts a two-stage generation pipeline: a coarse global structure is first synthesized and then refined to produce detailed, high-quality geometry. To support reliable 3D generation, we develop a comprehensive data processing pipeline that includes a novel watertight processing method and high-quality data filtering. This pipeline improves the geometric quality of publicly available 3D datasets by removing low-quality samples, filling holes, and thickening thin structures, while preserving fine-grained geometric details. To enable fine-grained geometry refinement, we decouple spatial localization from geometric detail synthesis in the diffusion process. We achieve this by performing voxel-based refinement at fixed spatial locations, where voxel queries derived from coarse geometry provide explicit positional anchors encoded via RoPE, allowing the diffusion model to focus on synthesizing local geometric details within a reduced, structured solution space. Our model is trained exclusively on publicly available 3D datasets, achieving strong geometric quality despite limited training resources. Extensive evaluations demonstrate that UltraShape 1.0 performs competitively with existing open-source methods in both data processing quality and geometry generation. All code and trained models will be released to support future research.