SGS-SLAM：面向神经稠密SLAM的语义高斯泼溅
SGS-SLAM: Semantic Gaussian Splatting For Neural Dense SLAM
AI总结：
SGS-SLAM是首个基于高斯泼溅的语义视觉SLAM系统，通过多通道优化融合外观、几何与语义特征，并引入语义特征损失和语义引导关键帧选择策略，在相机位姿估计、地图重建、语义分割及物体级几何精度上达到最先进水平，同时支持实时渲染。
摘要：
我们提出了SGS-SLAM，这是首个基于高斯泼溅（Gaussian Splatting）的语义视觉SLAM系统。该系统通过多通道优化融合外观、几何和语义特征，解决了神经隐式SLAM系统在高质量渲染、场景理解和物体级几何方面存在的过度平滑局限性。我们引入了一种独特的语义特征损失，有效弥补了传统深度损失和颜色损失在物体优化中的不足。通过语义引导的关键帧选择策略，我们防止了累积误差导致的错误重建。大量实验表明，SGS-SLAM在相机位姿估计、地图重建、精确语义分割和物体级几何精度方面均达到了最先进的性能，同时保证了实时渲染能力。
英文摘要：
We present SGS-SLAM, the first semantic visual SLAM system based on Gaussian Splatting. It incorporates appearance, geometry, and semantic features through multi-channel optimization, addressing the oversmoothing limitations of neural implicit SLAM systems in high-quality rendering, scene understanding, and object-level geometry. We introduce a unique semantic feature loss that effectively compensates for the shortcomings of traditional depth and color losses in object optimization. Through a semantic-guided keyframe selection strategy, we prevent erroneous reconstructions caused by cumulative errors. Extensive experiments demonstrate that SGS-SLAM delivers state-of-the-art performance in camera pose estimation, map reconstruction, precise semantic segmentation, and object-level geometric accuracy, while ensuring real-time rendering capabilities.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
