ObjectGS：基於高斯潑濺的物件感知場景重建與場景理解

摘要

3D高斯潑濺技術以其高保真重建與實時新視角合成而著稱，然而其缺乏語義理解能力，限制了物體層面的感知。在本研究中，我們提出了ObjectGS，這是一個物體感知框架，將3D場景重建與語義理解相統一。ObjectGS不再將場景視為一個整體，而是將各個物體建模為生成神經高斯並共享物體ID的局部錨點，從而實現精確的物體層面重建。在訓練過程中，我們動態地增長或修剪這些錨點，並優化其特徵，同時利用帶有分類損失的獨熱ID編碼來強化清晰的語義約束。通過大量實驗，我們展示了ObjectGS不僅在開放詞彙和全景分割任務上超越了現有最先進的方法，還能無縫集成於網格提取和場景編輯等應用中。項目頁面：https://ruijiezhu94.github.io/ObjectGS_page

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

ObjectGS：基於高斯潑濺的物件感知場景重建與場景理解

ObjectGS: Object-aware Scene Reconstruction and Scene Understanding via Gaussian Splatting

摘要

Support