扩散编辑中的可控性-保真度前沿
On the Controllability-Fidelity Frontier in Diffusion Editing
AI总结：
本文理论结合实证研究扩散图像编辑中用户意图遵循、非目标内容保持与输出质量间的权衡，提出算法框架并揭示关键失败模式，讨论伦理考量。
摘要：
基于扩散的生成模型实现了强大的图像编辑能力，但在保持保真度和安全性的同时实现精确控制仍然具有挑战性。我们对可控的基于扩散的图像编辑进行了全面的理论和实证研究，分析了用户意图遵循、非目标内容保持和输出质量之间的权衡。我们的工作涵盖了文本和掩码引导编辑、点/拖拽操作以及基于反演的流程。我们推导了编辑目标的数学公式，并分析了噪声注入、分数引导和反演误差的动力学。我们提供了重构误差、重复编辑下的稳定性以及变化局部性的理论界限。我们提出了掩码局部化和指令引导编辑的算法框架（附伪代码），并在多个任务和指标（FID、身份相似性、CLIP对齐、伪影分数等）上进行了广泛的实验，比较了最先进的方法（例如TF-ICON \cite{lu2023tficone}、DragFlow \cite{zhou2025dragflow}、InstructPix2Pix \cite{brooks2023instructpix2pix}、UltraEdit \cite{zhao2024ultraedit}）。我们的结果揭示了关键失败模式，如身份漂移、提示敏感性和组合错误。我们还讨论了图像编辑中的伦理考量，包括滥用风险、偏见、同意以及概念擦除技术（例如MACE \cite{lu2024mace}、ANT \cite{li2025ant}、EraseAnything \cite{gao2024eraseanything}）作为安全措施。最后，我们总结了负责任、高保真度扩散编辑的最佳实践和未来方向。
英文摘要：
Diffusion-based generative models enable powerful image editing capabilities, but achieving precise control while maintaining fidelity and safety remains challenging. We present a comprehensive theoretical and empirical study of controllable diffusion-based image editing, analyzing the trade-offs between adherence to user intent, preservation of non-target content, and output quality. Our work spans text- and mask-guided edits, point/drag manipulation, and inversion-based pipelines. We derive mathematical formulations of editing objectives and analyze dynamics of noise injection, score guidance, and inversion error. We provide theoretical bounds on reconstruction error, stability under repeated edits, and locality of changes. We propose algorithmic frameworks (with pseudocode) for mask-localized and instruction-guided editing, and present extensive experiments comparing state-of-the-art methods (e.g.\ TF-ICON \cite{lu2023tficone}, DragFlow \cite{zhou2025dragflow}, InstructPix2Pix \cite{brooks2023instructpix2pix}, UltraEdit \cite{zhao2024ultraedit}) on multiple tasks and metrics (FID, identity similarity, CLIP alignment, artifact scores, etc). Our results reveal key failure modes, such as identity drift, prompt sensitivity, and compositional errors. We also discuss ethical considerations in image editing, including misuse risks, bias, consent, and concept erasure techniques (e.g.\ MACE \cite{lu2024mace}, ANT \cite{li2025ant}, EraseAnything \cite{gao2024eraseanything}) as safeguards. We conclude with best practices and future directions for responsible, high-fidelity diffusion-based editing.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
