GSplatLoc：基于3D高斯溅射的超精确相机定位
GSplatLoc: Ultra-Precise Camera Localization via 3D Gaussian Splatting
AI总结：
本研究提出GSplatLoc相机定位方法，通过3D高斯溅射可微渲染将位姿估计转化为梯度优化问题，在Replica等数据集上实现远超现有方法的精度，为稠密建图定位设立新基准，可应用于机器人与AR领域。
摘要：
我们提出了GSplatLoc，这是一种利用3D高斯溅射的可微渲染能力实现超精确位姿估计的相机定位方法。通过将位姿估计建模为基于梯度的优化问题，最小化已有3D高斯场景渲染深度图与观测深度图像之间的差异，GSplatLoc在Replica数据集上实现了0.01 cm以内的平移误差和接近零的旋转误差，显著优于现有方法。在Replica和TUM RGB-D数据集上的评估表明，该方法在具有复杂相机运动的挑战性室内环境中具有鲁棒性。GSplatLoc为稠密建图中的定位设定了新基准，对机器人、增强现实等需要精确实时定位的应用具有重要意义。
英文摘要：
We present GSplatLoc, a camera localization method that leverages the differentiable rendering capabilities of 3D Gaussian splatting for ultra-precise pose estimation. By formulating pose estimation as a gradient-based optimization problem that minimizes discrepancies between rendered depth maps from a pre-existing 3D Gaussian scene and observed depth images, GSplatLoc achieves translational errors within 0.01 cm and near-zero rotational errors on the Replica dataset - significantly outperforming existing methods. Evaluations on the Replica and TUM RGB-D datasets demonstrate the method's robustness in challenging indoor environments with complex camera motions. GSplatLoc sets a new benchmark for localization in dense mapping, with important implications for applications requiring accurate real-time localization, such as robotics and augmented reality.
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。
