KFS$^4$LAM: Adaptive KeyFrame Selection with Structure-Aware Mapping for Endoscopic Gaussian Splatting SLAM

Gaussian-splatting Simultaneous Localization and Mapping (SLAM) augments conventional SLAM with a dense, renderable 3D scene representation, enabling high-quality reconstruction and novel-view synthesis. However, its localization, reconstruction, and rendering performance depends strongly on which observations are retained for mapping.

高斯溅射同步定位与建图（SLAM）通过密集的、可渲染的 3D 场景表示增强了传统 SLAM，从而实现了高质量的重建和新视角合成。然而，其定位、重建和渲染性能在很大程度上取决于哪些观测数据被保留用于建图。

Existing systems commonly rely on fixed temporal intervals or simple pose- and overlap-based criteria for keyframe selection. In colonoscopy, non-uniform camera motion, repetitive anatomy, specularities, and visual artifacts can make these criteria unreliable, leading to either redundant keyframes or missed informative viewpoints.

现有系统通常依赖固定的时间间隔或简单的基于位姿和重叠的标准来进行关键帧选择。在结肠镜检查中，非均匀的相机运动、重复的解剖结构、镜面反射和视觉伪影可能导致这些标准不可靠，从而导致关键帧冗余或错失具有信息量的视角。

We present KFS$^4$LAM (KeyFrame Selection with Structure-aware mapping for endoscopic Gaussian Splatting SLAM), which couples motion-guided adaptive keyframe selection with structurally supervised mapping and restricted temporal deformation. Accumulated motion guides adaptive keyframe selection, with complementary cues suppressing redundant or unreliable candidates.

我们提出了 KFS$^4$LAM（用于内窥镜高斯溅射 SLAM 的结构感知映射关键帧选择），它将运动引导的自适应关键帧选择与结构监督映射及受限的时间变形相结合。累积运动引导自适应关键帧选择，同时利用补充线索抑制冗余或不可靠的候选帧。

Density-aware mapping with anatomical contour supervision (colon folds) emphasizes structurally informative regions, while restricted temporal deformation accommodates local tissue dynamics without unconstrained geometric drift.

具有解剖轮廓监督（结肠褶皱）的密度感知映射强调了结构上具有信息量的区域，而受限的时间变形则在不产生无约束几何漂移的情况下适应了局部组织动态。

We evaluate KFS$^4$LAM against Gaussian-SLAM baselines on C3VD and C3VD-v2, together with cross-backbone and component ablations. KFS$^4$LAM achieves strong overall performance across trajectory, rendering, and depth metrics, with pronounced gains on more challenging C3VD-v2 dataset.

我们在 C3VD 和 C3VD-v2 数据集上评估了 KFS$^4$LAM 与高斯-SLAM 基线的对比情况，并进行了跨骨干网络和组件消融实验。KFS$^4$LAM 在轨迹、渲染和深度指标上均表现出强大的整体性能，在更具挑战性的 C3VD-v2 数据集上增益显著。

In the matched comparison with SplaTAM on C3VD-v2, it reduces rotational relative pose error by 20.3% and depth RMSE by 38.4%, while improving PSNR by 0.78 dB. Cross-backbone experiments further show that the framework is not tied to a single underlying SLAM architecture.

在与 C3VD-v2 上 SplaTAM 的匹配比较中，它将旋转相对位姿误差降低了 20.3%，深度 RMSE 降低了 38.4%，同时将 PSNR 提高了 0.78 dB。跨骨干网络实验进一步表明，该框架并不绑定于单一的底层 SLAM 架构。