[论文] HiP3R: Hierarchical Pose Graphs for Long-Horizon 3D Reconstruction

论文概要

研究领域: CV
作者: Zunnan Xu, Shangchen Zhou, Yihang Luo, Chen Change Loy
发布时间: 2026-10-22
arXiv: 2610.31635

中文摘要

前馈模型能从短视频窗口中准确估计相机运动和3D几何，但计算和内存成本阻碍了对包含数千帧序列的直接推理。长视频必须被切分为重叠块处理，这产生了坐标系不一致、共享相机重复估计的独立局部重建。HiP3R用层次化位姿图解决这一权衡：首先优化鲁棒的块级Sim(3)图建立全局布局和尺度，然后固定布局为每帧估计共享SE(3)相机，将重叠窗口预测转换为单一轨迹。多个基准实验证明HiP3R有效改善相机轨迹和3D重建，且无需额外训练或微调即可适用于基于块和流式后端。

原文摘要

Feed-forward models accurately estimate camera motion and 3D geometry from short video windows. However, computational and memory costs prevent direct inference on sequences containing thousands of frames. Long videos must therefore be processed as overlapping chunks. This division produces independent local reconstructions with inconsistent coordinate frames and duplicate estimates of shared cameras. Achieving global consistency requires propagating constraints from adjacent overlaps and distant revisits across the sequence. This requirement naturally motivates pose graph optimization. However, the choice of graph level creates a trade-off. Chunk-level graphs efficiently resolve scale drift and close large loops. Yet, their rigid transformations cannot correct internal errors or conflicti...

自动采集于 2026-10-26

#论文 #arXiv #CV #小凯