TripleFlow：通过桥接残差编辑与原生生成实现免训练视频物体移除
TripleFlow: Training-Free Video Object Removal by Bridging Residual Editing and Native Generation
浏览论文内容
中文总结 AI 辅助
TripleFlow通过协调源流、残差流和合成流，在每一步注入新合成背景，实现免训练视频物体移除，在五个基准上显著优于现有方法。
中文摘要 AI 辅助
视频物体移除是一项极具挑战性的编辑任务。由于移除提示仅指定要擦除的内容，而非要生成的内容，模型必须完全根据周围上下文推断并重建高度特定的遮挡背景。现有的免训练方法在此任务中表现不佳，因为其编辑机制主要充当局部擦除器，无法主动合成缺失的背景细节，且常常留下重影伪影。为解决这一问题，我们提出了TripleFlow，一个免训练框架，紧密耦合擦除与生成。它协调源流、残差流和合成流贯穿整个过程。通过复用单一目标预测，残差流隔离并抑制物体，而合成流独立重建被遮挡的背景。关键在于，TripleFlow在每一步都将新合成的背景注入编辑轨迹中。这种持续反馈循环确保生成的结构主动引导移除过程，实现与未编辑场景时空一致的无缝完成。在五个具有挑战性的基准上的广泛评估表明，TripleFlow确立了新的最先进水平，在重建保真度和时间一致性方面显著优于现有基线。
英文摘要
Video object removal presents a uniquely difficult editing challenge. Because a removal prompt specifies only what to erase rather than what to generate, the model must infer and reconstruct a highly specific occluded background entirely from the surrounding context. Existing training-free methods struggle with this because their editing mechanisms act primarily as localized erasers. They fail to actively synthesize the missing background details and often leave behind ghosting artifacts. To solve this, we propose TripleFlow, a training-free framework that tightly couples erasure and generation. It coordinates a source flow, a residual flow, and a synthesis flow throughout the entire process. By reusing a single target prediction, the residual flow isolates and suppresses the object, while the synthesis flow independently reconstructs the occluded background. Crucially, TripleFlow injects this newly synthesized background back into the editing trajectory at every step. This continuous feedback loop ensures that the generated structures actively guide the removal process, achieving seamless completion that is spatiotemporally consistent with the unedited scene. Extensive evaluations across five challenging benchmarks demonstrate that TripleFlow establishes a new state-of-the-art, significantly outperforming existing baselines in both reconstruction fidelity and temporal consistency.
发表机构
- Penn State University（宾夕法尼亚州立大学）
- The University of British Columbia（不列颠哥伦比亚大学）
- Bournemouth University（伯恩茅斯大学）
- DEVCOM Army Research Laboratory（DEVCOM陆军研究实验室）
机构由 AI 辅助整理，请以论文原文为准。
