用SAM-HQ和多视角一致性实现高精度3D场景分割，支持实时编辑。
Robust Prior-Guided Segmentation for Editable 3D Gaussian Splatting
- 结合SAM-HQ生成精细2D掩码，提升边界保真度。
- 通过先验引导的标签重分配实现多视角一致的3D分割。
- 支持虚拟现实与机器人应用中的交互式实时编辑。
3D Gaussian Splatting（3D-GS）实现了实时3D场景重建，但在物体移除、提取和着色等编辑任务中缺乏鲁棒的分割能力。现有方法将2D分割映射到3D域时存在视角不一致和掩码粗糙的问题。本文提出新框架，利用高精度的Segment Anything Model High Quality（SAM-HQ）生成准确的2D掩码，解决标准SAM在边界保真度和细结构保留上的不足。为实现场景中任意目标对象的鲁棒3D分割，引入先验引导的标签重分配方法，通过学习的先验强制多视角一致性地为3D高斯分布分配标签。本方法在分割精度上达到当前最优，并支持交互式、实时的对象编辑，同时保持高视觉保真度。定性结果表明，其在边界保留方面表现优异，适用于虚拟现实（VR）与机器人领域，推动了3D场景编辑的发展。
原文摘要 · Abstract (English)
3D Gaussian Splatting (3D-GS) enables real-time 3D scene reconstruction but lacks robust segmentation for editing tasks such as object removal, extraction, and recoloring. Existing approaches that lift 2D segmentations to the 3D domain suffer from view inconsistencies and coarse masks. In this paper, we propose a novel framework that leverages the Segment Anything Model High Quality (SAM-HQ) to generate accurate 2D masks, addressing the limitations of the standard SAM in boundary fidelity and fine-structure preservation. To achieve robust 3D segmentation of any target object in a given scene, we introduce a prior-guided label reassignment method that assigns labels to 3D Gaussians by enforcing multiview consistency with learned priors. Our approach achieves state-of-the-art segmentation accuracy and enables interactive, real-time object editing while maintaining high visual fidelity. Qualitative results demonstrate superior boundary preservation and practical utility in Virtual Reality (VR) and robotics, advancing 3D scene editing.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。