GaussianSelector：基于图优化的轻量级 3D 高斯泼溅对象选择
原标题：GaussianSelector: Lightweight Human-Guided Object Selection in 3D Gaussian Splatting with Graph Optimization
AI 摘要
Hugging Face 每日论文发布了一篇题为《GaussianSelector: Lightweight Human-Guided Object Selection in 3D Gaussian Splatting with Graph Optimization》的论文。该论文提出了一种名为 GaussianSelector 的无需训练框架，用于在 3D 高斯泼溅场景中通过稀疏视图和稀疏涂鸦进行交互式对象选择。该方法通过将高斯原语粗化为超点并构建图，将用户涂鸦提升为 3D 选择，并通过图割优化实现完整对象分割，支持多轮细化，计算开销低，适合实际部署。
正文节选
GaussianSelector: Lightweight Human-Guided Object Selection in 3D Gaussian Splatting with Graph Optimization Abstract Selecting a complete 3D object from a reconstructed scene with minimal user effort is essential for practical scene editing and embodied interaction. Existing 3DGS-based methods either retrain the Gaussian representation to embed per-object labels, or build dense multi-view SAM observations, both requiring heavy computation and dense viewpoint coverage that is rarely available in