FullPart: Generating each 3D Part at Full Resolution

2025-10-30

AI Summary

Existing part-based 3D generation methods face a fundamental trade-off: implicit representations (e.g., point sets) lack geometric fidelity, while explicit voxel representations suffer from distortion of small parts due to shared low-resolution global grids. This work proposes an implicit–explicit collaborative framework: first, an implicit vector-set diffusion model generates coarse part bounding-box layouts; then, each part is assigned a dedicated high-resolution voxel grid, and a center-point encoding scheme ensures precise spatial alignment across multiple parts. To support training and evaluation, we introduce PartVerse-XL—a large-scale, fine-grained part-level 3D dataset. Experiments demonstrate state-of-the-art performance in geometric fidelity, small-part quality, and global structural consistency. The code, pretrained models, and dataset will be publicly released.

Technology Category

• Computer Vision: 3D Computer Vision
• Machine Learning: Deep Generative Models & Autoencoders
• Knowledge Representation and Reasoning: Geometric, Spatial, and Temporal Reasoning

Application Category

• Semantics and Knowledge: Scalable techniques for the creation, curation, publication, maintenance, and consumption of large, Web-based, structured, reusable, knowledge graphs and ontologies
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets

Abstract

Part-based 3D generation holds great potential for various applications. Previous part generators that represent parts using implicit vector-set tokens often suffer from insufficient geometric details. Another line of work adopts an explicit voxel representation but shares a global voxel grid among all parts; this often causes small parts to occupy too few voxels, leading to degraded quality. In this paper, we propose FullPart, a novel framework that combines both implicit and explicit paradigms. It first derives the bounding box layout through an implicit box vector-set diffusion process, a task that implicit diffusion handles effectively since box tokens contain little geometric detail. Then, it generates detailed parts, each within its own fixed full-resolution voxel grid. Instead of sharing a global low-resolution space, each part in our method - even small ones - is generated at full resolution, enabling the synthesis of intricate details. We further introduce a center-point encoding strategy to address the misalignment issue when exchanging information between parts of different actual sizes, thereby maintaining global coherence. Moreover, to tackle the scarcity of reliable part data, we present PartVerse-XL, the largest human-annotated 3D part dataset to date with 40K objects and 320K parts. Extensive experiments demonstrate that FullPart achieves state-of-the-art results in 3D part generation. We will release all code, data, and model to benefit future research in 3D part generation.

Problem

• Generating detailed 3D parts at full resolution
• Addressing insufficient geometric details in part generation
• Solving misalignment between parts of different sizes

Innovation

• Combines implicit and explicit 3D generation paradigms
• Generates each part in fixed full-resolution voxel grid
• Uses center-point encoding for inter-part alignment