VDFE：基于非侵入式视频扩散先验的差异感知3D场景编辑，实现多视图一致性与高效性
VDFE: Difference-Aware 3D Scene Editing with Non-Intrusive Video Diffusion Priors for Multi-View Consistency and Efficiency
将视频扩散模型作为非侵入式先验，通过最优控制优化无噪声编辑轨迹，并利用流差异精确定位编辑区域，为3D场景编辑提供多视图一致且高效的新思路。
原文依据 we propose VDFE, a difference-aware 3D scene editing method based on non-intrusive utilization of pre-trained video diffusion priors
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论文本驱动的3D编辑借助神经辐射场和3D高斯泼溅等3D重建技术的进步，旨在提供直观的场景定制。然而，现有方法在可控性和一致性方面常表现出局限。为解决这些不足，我们提出VDFE，一种基于非侵入式利用预训练视频扩散先验的差异感知3D场景编辑方法，它集成了最优控制引导的流编辑（FlowOCE）、解耦流差异（DFD）和差异感知高斯编辑（DAGE）模块。具体而言，FlowOCE将编辑过程视为最优控制问题，优化无噪声的编辑轨迹以最小化非目标区域的意外修改，并生成具有多视图一致性和平滑过渡的结果；DFD通过分析流差异，创新性地实现了比交叉注意力机制更精确的编辑区域定位，为后续优化过程提供先验；DAGE利用精确定位选择性地更新3D高斯，实现高效且精确的细化。大量实验表明，我们的方法在定性和定量评估中均显著优于现有方法，达到了最优性能。
英文原文摘要 1340 字符
Text-driven 3D editing, enabled by advancements in 3D reconstruction techniques such as NeRF and 3D Gaussian Splatting, aims to provide intuitive scene customization. However, existing methods frequently exhibit limitations in controllability and consistency. To address these shortcomings, we propose VDFE, a difference-aware 3D scene editing method based on non-intrusive utilization of pre-trained video diffusion priors, which integrates Optimal Control Guided Flow Editing (FlowOCE), Decoupled Flow Difference (DFD), and Difference-Aware Gaussians Editing (DAGE) modules. Specifically, FlowOCE treats the editing process as an optimal control problem, optimizing a noise-free editing trajectory to minimize unintended modifications in non-target regions and generating results with multi-view consistency and smooth transitions; DFD innovatively achieves more precise localization of editing regions compared to cross-attention mechanisms by analyzing flow differences, which supplies priors for the subsequent optimization process; and DAGE leverages precise localization to selectively update 3D Gaussians for efficient and precise refinement. Extensive experiments demonstrate that our method significantly outperforms existing methods in both qualitative and quantitative evaluations, achieving state-of-the-art (SOTA) performance.