让3D高斯点阵具备物体级语义理解，实现精准重建与场景编辑。
ObjectGS: Object-aware Scene Reconstruction and Scene Understanding via Gaussian Splatting
- 以物体为局部锚点生成高斯点，共享对象ID实现语义对齐
- 动态增删锚点并优化特征，提升开放词汇与全景分割性能
- 适用于网格提取与场景编辑，兼具精度与实用性
3D高斯点阵以高保真重建和实时新视角合成著称，但缺乏语义理解，限制了物体级感知。本文提出ObjectGS，一个统一3D场景重建与语义理解的物体感知框架。不同于将场景视为整体，ObjectGS将每个物体建模为局部锚点，生成神经高斯点并共享对象ID，实现精确的物体级重建。训练中，动态增长或剪枝这些锚点并优化其特征，同时使用独热编码与分类损失强化语义约束。大量实验表明，ObjectGS在开放词汇和全景分割任务上均超越现有方法，并可无缝集成至网格提取与场景编辑应用中。
原文摘要 · Abstract (English)
3D Gaussian Splatting is renowned for its high-fidelity reconstructions and real-time novel view synthesis, yet its lack of semantic understanding limits object-level perception. In this work, we propose ObjectGS, an object-aware framework that unifies 3D scene reconstruction with semantic understanding. Instead of treating the scene as a unified whole, ObjectGS models individual objects as local anchors that generate neural Gaussians and share object IDs, enabling precise object-level reconstruction. During training, we dynamically grow or prune these anchors and optimize their features, while a one-hot ID encoding with a classification loss enforces clear semantic constraints. We show through extensive experiments that ObjectGS not only outperforms state-of-the-art methods on open-vocabulary and panoptic segmentation tasks, but also integrates seamlessly with applications like mesh extraction and scene editing. Project page: https://ruijiezhu94.github.io/ObjectGS_page
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。