并非所有时间和频率都需要在扩散遗忘中被遗忘
Not Every Time and Frequency Need to Be Forgotten in Diffusion Unlearning
AI总结：
提出选择性遗忘扩散模型中的时间和频率，通过理论分析分布失真与遗忘-效用权衡，在多种设置下实现更高遗忘成功率和生成质量。
摘要：
数据遗忘旨在从训练模型中移除特定训练样本的影响。在微调方法中，数据遗忘主要依赖于对遗忘样本的损失最大化，这常常导致质量下降或遗忘不彻底。现有方法在扩散阶段统一进行遗忘，忽略了从噪声到数据的扩散动态。我们对扩散阶段的系统研究表明，扩散模型中的遗忘在时间和频率上是不均匀的，并给出了分布失真和遗忘-效用权衡的理论证明。通过选择性地遗忘扩散模型中的时间和频率，我们在包括条件和无条件场景在内的多种设置下，同时实现了更高的遗忘成功率和改进的生成质量。我们还引入了一种改进的SSCD指标，该指标使用归一化扰动距离来衡量不相似性。总之，我们为理解和改进扩散模型中的数据遗忘提供了实用见解。
英文摘要：
Data unlearning aims to remove the influence of specific training samples from a trained model. In fine-tuning methods, data unlearning relies primarily on loss maximization over forget samples, which often leads to quality degradation or incomplete forgetting. Existing methods perform unlearning uniformly across diffusion stages, ignoring diffusion dynamics from noise to data. Our systematic study of diffusion phases shows that forgetting in diffusion models is uneven across time and frequency, with theoretical justification of distributive distortion and forgetting-utility trade-off. By selectively forgetting time and frequency in diffusion models, we achieve both higher unlearning success rates and improved generation quality across diverse settings, including both conditional and unconditional scenarios. We also introduce an improved SSCD metric that measures dissimilarity using a normalized perturbation distance. Together, we provide practical insights for understanding and improving data unlearning in diffusion models.
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。
