发表机构
University of Illinois Urbana-Champaign（伊利诺伊大学厄巴纳-香槟分校）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
针对单一框架内实现高质量多样化视频编辑的挑战，提出无训练框架EditVid，结合稀疏因果记忆等技术，在FiVE、IVEBench数据集及用户研究中均取得优于基线的结果。
AI 中文摘要
视频编辑涵盖多样化的编辑范式，但在单一统一框架内实现高质量的指令引导和主体引导编辑仍具挑战性。我们提出EditVid，这是一种无训练框架，结合了用于局部连贯性的稀疏因果记忆、用于长程身份保留的基于对应关系的后注意力令牌注入，以及用于编辑局部性的软潜在混合。该相同框架支持指令引导和参考引导编辑，包括风格迁移、属性修改、对象插入、部件级编辑和主体替换。在FiVE数据集上，EditVid达到78.16的FiVE-Acc，相比之下，评估中最强的无训练基线为58.95；在IVEBench上也获得了具有竞争力的结果。用户研究进一步显示，与7种对比方法相比，EditVid的总体偏好率为51.8%。
英文摘要
Video editing spans diverse editing paradigms, yet achieving high-quality instruction-guided and subject-guided editing within a single unified framework remains challenging. We introduce EditVid, a training-free framework combining sparse causal memory for local coherence, correspondence-based post-attention token injection for long-range identity preservation, and soft latent blending for edit locality. The same framework supports instruction-guided and reference-guided edits, including style transfer, attribute modification, object insertion, part-level editing, and subject replacement. On FiVE, EditVid achieves 78.16 FiVE-Acc, compared with 58.95 for the strongest evaluated training-free baseline, while obtaining competitive results on IVEBench. A user study further shows a 51.8\% overall preference for EditVid over 7 competing methods.
Commentshttps://plan-lab.github.io/editvid
