ClearGS：从手持视频重建可靠 3D Gaussian Splatting
ClearGS: Reliability-Aware Gaussian Splatting from Handheld Videos
手持视频拍糊了也能重建 3D 场景，这篇 ClearGS 靠加权监督加无参考修复，在 GS2E 和 GSOTM 上跑赢了现有方法。
ClearGS 是一个面向手持视频的 3D Gaussian Splatting 方法，针对视角覆盖不均和帧质量混杂的问题。它用 Reliability-aware View Allocation（RVA）按外观可靠性、退化风险和几何效用分配分级监督权重，而非二值选帧。针对模糊导致的信息丢失，Render-Guided In-Video Restoration（RIVR）借助冻结的无参考修复专家恢复原始视频帧，无需干净参考图。在 GS2E 和 GSOTM 两个基准上，ClearGS 取得整体最优表现，CLIP-IQA 和 MUSIQ 一致提升，多数退化设置下 LPIPS 下降，且不依赖成对清晰监督。
ClearGS: Reliability-Aware Gaussian Splatting from Handheld Videos
We present ClearGS for 3D Gaussian Splatting (3DGS) from handheld videos with uneven viewpoint coverage and mixed frame quality. Rather than selecting frames with binary decisions, ClearGS uses Reliability-aware View Allocation (RVA) to assign graded raw-supervision weights based on appearance reliability, degradation risk, and geometric utility, while weakly reactivating useful suppressed frames to maintain trajectory coverage. Since weighting cannot restore details lost to blur or distortion, ClearGS further introduces Render-Guided In-Video Restoration (RIVR). The current 3DGS render provides a pose-aligned structural candidate, a frozen no-reference restoration expert restores the corresponding raw video observation without any clean reference image, and no-reference perceptual scores select among the render, restored observation, and high-frequency fused candidate. ClearGS then applies Full-Trajectory Repair Consolidation to revisit accepted repairs and preserve details introduced early. On GS2E and GSOTM, ClearGS achieves state-of-the-art overall performance, with consistent CLIP-IQA and MUSIQ gains and LPIPS reductions in most degradation settings, without paired sharp supervision or matched clean references.
