MotionEdit：以运动为中心的图像编辑基准与学习方法
MotionEdit: Benchmarking and Learning Motion-Centric Image Editing
MotionNFT 提出了一种基于运动流对齐奖励的扩散模型后训练微调框架，可复用于其他需要精确运动控制的图像编辑任务，如帧控制视频合成中的关键帧编辑。
原文依据 we propose MotionNFT (Motion-guided Negative-aware FineTuning), a post-training framework that computes motion alignment rewards based on how well the motion flow between input and model-edited images matches the ground-truth motion
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论我们提出了 MotionEdit，一个用于以运动为中心的图像编辑的新数据集——该任务旨在修改主体动作和交互，同时保持身份、结构和物理合理性。与现有专注于静态外观变化或仅包含稀疏、低质量运动编辑的图像编辑数据集不同，MotionEdit 提供了高保真图像对，描绘了从连续视频中提取并验证的真实运动变换。这一新任务不仅具有科学挑战性，还具有实际重要性，可推动帧控制视频合成和动画等下游应用。为了评估模型在新任务上的性能，我们推出了 MotionEdit-Bench，一个在运动中心编辑上挑战模型并采用生成式、判别式和基于偏好的指标来衡量模型性能的基准。基准结果表明，运动编辑对现有最优的基于扩散的编辑模型仍然极具挑战性。为弥补这一差距，我们提出了 MotionNFT（运动引导的负感知微调），一个后训练框架，它根据输入图像与模型编辑图像之间的运动流与真实运动的匹配程度来计算运动对齐奖励，引导模型实现准确的运动变换。在 FLUX.1 Kontext 和 Qwen-Image-Edit 上的大量实验表明，MotionNFT 在不牺牲通用编辑能力的情况下，持续提升了两个基础模型在运动编辑任务上的编辑质量和运动保真度，证明了其有效性。
英文原文摘要 1582 字符
We introduce **MotionEdit**, a novel dataset for motion-centric image editing--the task of modifying subject actions and interactions while preserving identity, structure, and physical plausibility.Unlike existing image editing datasets that focus on static appearance changes or contain only sparse, low-quality motion edits, MotionEdit provides high-fidelity image pairs depicting realistic motion transformations extracted and verified from continuous videos. This new task is not only scientifically challenging but also practically significant, powering downstream applications such as frame-controlled video synthesis and animation.To evaluate model performance on the novel task, we introduce **MotionEdit-Bench**, a benchmark that challenges models on motion-centric edits and measures model performance with generative, discriminative, and preference-based metrics.Benchmark results reveal that motion editing remains highly challenging for existing state-of-the-art diffusion-based editing models.To address this gap, we propose **MotionNFT** (Motion-guided Negative-aware FineTuning), a post-training framework that computes motion alignment rewards based on how well the motion flow between input and model-edited images matches the ground-truth motion, guiding models toward accurate motion transformations.Extensive experiments on FLUX.1 Kontext and Qwen-Image-Edit show that MotionNFT consistently improves editing quality and motion fidelity of both base models on the motion editing task without sacrificing general editing ability, demonstrating its effectiveness.