用树形结构统一建模动态世界时空关系，提升单目视频重建精度
WorldTree: Towards 4D Dynamic Worlds from Monocular Video using Tree-Chains
- 提出时序树与空间祖先链的分层框架，解耦时空建模
- 在NVIDIA-LS和DyCheck数据集上分别提升8.26%和9.09%的LPIPS指标
- 适合关注单目动态场景重建的视觉算法研究者
动态重建已取得显著进展，但在单目输入下仍面临挑战。现有方法虽尝试构建高效运动表征，但缺乏统一的时空分解框架，或依赖整体时序优化，或存在层级空间结构耦合问题。为此，我们提出WorldTree，包含时序分割树（TPT）与空间祖先链（SAC）：TPT基于继承式树结构实现粗粒度到细粒度的分层时序优化；SAC通过递归查询祖先层级结构，为各节点提供互补的空间动态信息，并实现运动表征的层次化特化。在多个数据集上的实验表明，该方法在NVIDIA-LS上相较第二佳方法提升8.26% LPIPS，于DyCheck上提升9.09% mLPIPS。代码已开源。
原文摘要 · Abstract (English)
Dynamic reconstruction has achieved remarkable progress, but there remain challenges in monocular input for more practical applications. The prevailing works attempt to construct efficient motion representations, but lack a unified spatiotemporal decomposition framework, suffering from either holistic temporal optimization or coupled hierarchical spatial composition. To this end, we propose WorldTree, a unified framework comprising Temporal Partition Tree (TPT) that enables coarse-to-fine optimization based on the inheritance-based partition tree structure for hierarchical temporal decomposition, and Spatial Ancestral Chains (SAC) that recursively query ancestral hierarchical structure to provide complementary spatial dynamics while specializing motion representations across ancestral nodes. Experimental results on different datasets indicate that our proposed method achieves 8.26% improvement of LPIPS on NVIDIA-LS and 9.09% improvement of mLPIPS on DyCheck compared to the second-best method. Code: https://github.com/iCVTEAM/WorldTree.
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。