Unblur-SLAM：面向模糊输入的稠密神经 SLAM
Unblur-SLAM: Dense Neural SLAM for Blurry Inputs
将前馈去模糊网络与 3DGS 建图耦合：成功去模糊的帧走局部-全局多视图优化与回环，失败帧则用额外模糊网络在 3D 空间显式建模模糊子帧形成过程，为模糊输入下的位姿估计与清晰重建提供了可复用的两阶段设计。
原文依据 Frames that fail the first stage deblurring are directly modeled through the global 3DGS representation and an additional blur network
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论我们提出 Unblur-SLAM，一种从模糊图像输入中实现清晰三维重建的 RGB SLAM 流程。与以往工作不同，我们的方法能够处理不同类型的模糊，并在同时存在运动模糊与离焦模糊的情况下展现出最优性能。此外，我们根据输入图像中模糊的程度来调整计算量。在第一阶段，我们的方法使用一个前馈图像去模糊模型，并为其提出了合适的训练方案，该方案能够同时改进跟踪与建图模块。被前馈网络成功去模糊的帧通过局部-全局多视图优化与回环检测获得精化的位姿与深度。未能通过第一阶段去模糊的帧则直接通过全局 3DGS 表示与一个额外的模糊网络进行建模，以建模多个模糊子帧并在三维空间中模拟模糊形成过程，从而学习清晰细节与精化的子帧位姿。在多个真实世界数据集上的实验表明，该方法在姿态估计以及几何与纹理的清晰重建结果上均取得了一致的提升。
英文原文摘要 1147 字符
We propose Unblur-SLAM, an RGB SLAM pipeline for sharp 3D reconstruction from blurred image inputs. In contrast to previous work, our approach is able to handle different types of blur and demonstrates state-of-the-art performance in the presence of both motion blur and defocus blur. Moreover, we adjust the computation effort with the amount of blur in the input image.As a first stage, our method uses a feed-forward image deblurring model for which we propose a suitable training scheme that can improve both tracking and mapping modules.Frames that are successfully deblurred by the feed-forward network obtain refined poses and depth through local-global multi-view optimization and loop closure. Frames that fail the first stage deblurring are directly modeled through the global 3DGS representation and an additional blur network to model multiple blurred sub-frames and simulate the blur formation process in 3D space, thereby learning sharp details and refined sub-frame poses.Experiments on several real-world datasets demonstrate consistent improvements in both pose estimation and sharp reconstruction results of geometry and texture.