突破单目视频动态重建长度限制，实现长时序一致的3D场景重建。
LongDPM: Overlap-Aware 4D Reconstruction from Long Monocular Videos
- 分块处理长视频并利用重叠区域对齐局部坐标系
- 在PointOdyssey等数据集上显著降低稠密追踪误差
- 适合需要长时间动态3D重建的自动驾驶与VR应用
从长序列单目视频中恢复动态3D场景对于保持密集几何、相机运动和时间对应关系的一致性至关重要。现有方法面临两大挑战：(1) 前馈重建模型虽能提供精确的局部预测，但仅适用于短片段；(2) 长程追踪器可维持对应关系，但无法生成完整的序列级稠密重建。本文提出LongDPM，一种新型重叠感知的4D动态重建框架，支持可扩展的长时序单目重建。首先，LongDPM将长视频分割为重叠块，在块内进行推理，从而将内存占用控制在块长范围内。其次，通过置信度加权的注册与静态感知的重叠抽象，连接各块内的局部坐标系。第三，跨块关联动态身份并融合匹配轨迹，实现连贯的长时序3D运动恢复。实验表明，LongDPM在PointOdyssey、Kubric-F和Kubric-G数据集上相较V-DPM显著降低稠密追踪端点误差（EPE），并在TUM-dynamics数据集上取得最优的相机位姿估计绝对轨迹误差（ATE）。
原文摘要 · Abstract (English)
Recovering a dynamic 3D scene from a long monocular video is crucial for dense geometry, camera motion, and temporal correspondence to remain consistent in a shared coordinate system. Existing methods face two key challenges: (1) feed-forward reconstruction models provide accurate local predictions but are limited to short clips, and (2) long-range trackers preserve correspondences without producing dense sequence-level reconstruction. This paper presents LongDPM, a novel overlap-aware framework for scalable long-range monocular dynamic reconstruction. First, LongDPM processes long videos in overlapping chunks, keeping inference memory bounded by the chunk length. Second, it connects chunk-local coordinate systems through confidence-weighted registration with static-aware overlap abstraction. Third, it associates dynamic identities across chunk boundaries and fuses matched trajectories to recover coherent long-range 3D motion. Experimental results demonstrate that LongDPM achieves superior long-range reconstruction and tracking performance, reducing dense tracking EPE over V-DPM on PointOdyssey, Kubric-F, and Kubric-G, while obtaining the best TUM-dynamics ATE for camera pose estimation.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。