以高质量合成数据集扩展指令式视频编辑
Scaling Instruction-Based Video Editing with a High-Quality Synthetic Dataset
用「图像编辑器 + 上下文视频生成器 + 智能体过滤」的流水线批量合成百万级指令式视频编辑数据，并用蒸馏模型加时序增强器压低生成成本；做视频编辑或可控生成的人可以直接复用这套数据构造与课程学习训练范式。
原文依据 a novel data generation pipeline that fuses the creative diversity of a leading image editor with an in-context video generator
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论指令式视频编辑有望让内容创作大众化，但其进展因缺乏大规模、高质量的训练数据而严重受阻。我们提出 Ditto，一个旨在解决这一根本性挑战的整体框架。Ditto 的核心是一个新颖的数据生成流水线，它将领先图像编辑器的创意多样性与上下文视频生成器相融合，克服了现有模型适用范围有限的缺陷。为使该过程可行，我们的框架通过采用高效的蒸馏模型架构并辅以时序增强器，解决了成本与质量之间难以兼顾的权衡，在降低计算开销的同时提升了时序一致性。最后，为实现完全的可扩展性，整条流水线由一个智能体驱动，它构造多样化的指令并对输出进行严格过滤，从而在规模化条件下保证质量控制。借助该框架，我们投入超过 12000 GPU 天构建了 Ditto-1M，这是一个包含一百万条高保真视频编辑样例的新数据集。我们采用课程学习策略在 Ditto-1M 上训练了模型 Editto。结果表明，该模型具备更优的指令遵循能力，并在指令式视频编辑上确立了新的最优水平。
英文原文摘要 1284 字符
Instruction-based video editing promises to democratize content creation, yet its progress is severely hampered by the scarcity of large-scale, high-quality training data. We introduce Ditto, a holistic framework designed to tackle this fundamental challenge. At its heart, Ditto features a novel data generation pipeline that fuses the creative diversity of a leading image editor with an in-context video generator, overcoming the limited scope of existing models. To make this process viable, our framework resolves the prohibitive cost-quality trade-off by employing an efficient, distilled model architecture augmented by a temporal enhancer, which simultaneously reduces computational overhead and improves temporal coherence. Finally, to achieve full scalability, this entire pipeline is driven by an intelligent agent that crafts diverse instructions and rigorously filters the output, ensuring quality control at scale. Using this framework, we invested over 12,000 GPU-days to build Ditto-1M, a new dataset of one million high-fidelity video editing examples. We trained our model, Editto, on Ditto-1M with a curriculum learning strategy. The results demonstrate superior instruction-following ability and establish a new state-of-the-art in instruction-based video editing.