GarmentGS: Point-Cloud Guided Gaussian Splatting for High-Fidelity Non-Watertight 3D Garment Reconstruction

2025-05-04

AI Summary

Traditional 3D garment reconstruction relies heavily on manual modeling, resulting in low efficiency, while existing 3D Gaussian Splatting (GS) methods struggle with non-watertight, non-manifold geometries. To address these limitations, this work proposes a point-cloud-guided Gaussian Splatting framework, enabling high-fidelity reconstruction and real-time rendering of single-layer, non-watertight garment meshes for the first time. Our key contributions are: (1) dense point cloud guidance for motion, flattening, and orientation optimization of Gaussian primitives; (2) differentiable surface-guided sampling and adaptive Gaussian orientation control; and (3) integration of fast point cloud reconstruction with point-cloud-constrained Gaussian optimization. Experiments demonstrate that point cloud reconstruction completes in just 10 minutes—over an order of magnitude faster than conventional methods—while reducing geometric error by 32% and achieving real-time rendering performance with state-of-the-art visual quality.

Technology Category

• Computer Vision: 3D Computer Vision
• Planning, Routing, and Scheduling: Replanning and Plan Repair
• Search and Optimization: Non-convex Optimization

Application Category

• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation
• Security and Privacy: Data transparency and provenance
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs

Abstract

Traditional 3D garment creation requires extensive manual operations, resulting in time and labor costs. Recently, 3D Gaussian Splatting has achieved breakthrough progress in 3D scene reconstruction and rendering, attracting widespread attention and opening new pathways for 3D garment reconstruction. However, due to the unstructured and irregular nature of Gaussian primitives, it is difficult to reconstruct high-fidelity, non-watertight 3D garments. In this paper, we present GarmentGS, a dense point cloud-guided method that can reconstruct high-fidelity garment surfaces with high geometric accuracy and generate non-watertight, single-layer meshes. Our method introduces a fast dense point cloud reconstruction module that can complete garment point cloud reconstruction in 10 minutes, compared to traditional methods that require several hours. Furthermore, we use dense point clouds to guide the movement, flattening, and rotation of Gaussian primitives, enabling better distribution on the garment surface to achieve superior rendering effects and geometric accuracy. Through numerical and visual comparisons, our method achieves fast training and real-time rendering while maintaining competitive quality.

Problem

• Reconstruct high-fidelity non-watertight 3D garments
• Reduce manual effort in 3D garment creation
• Improve Gaussian primitive distribution for better rendering

Innovation

• Point-cloud guided Gaussian Splatting for garments
• Fast dense point cloud reconstruction module
• Gaussian primitives optimized for garment surfaces