SUGAR可批量移除3D生成模型中多人身份，无需重训练。
SUGAR: A Sweeter Spot for Generative Unlearning of Many Identities
- 为每个身份学习个性化潜空间，引导重建到合理替代输出。
- 可移除最多200人身份，保留性能提升达700%。
- 适合需隐私保护的生成模型应用，如人脸合成去标识化。
近年来，3D感知生成模型实现了对人类身份的高保真图像合成，但这也带来了用户同意和从模型输出中移除特定个体的紧迫问题。我们提出SUGAR框架，实现可扩展的生成式去标识，无需重新训练整个模型即可同时或顺序移除多个身份。不同于将不想要的身份投影到不现实的输出或依赖静态模板人脸，SUGAR为每个身份学习个性化的代理潜变量，将重建结果导向视觉上连贯的替代方案，同时保持模型的质量与多样性。我们进一步引入持续性效用保全目标，防止随着更多身份被遗忘而性能退化。SUGAR在移除最多200个身份时达到当前最优表现，相比现有基线在保留效用上提升高达700%。代码已公开于https://github.com/judydnguyen/SUGAR-Generative-Unlearn。
原文摘要 · Abstract (English)
Recent advances in 3D-aware generative models have enabled high-fidelity image synthesis of human identities. However, this progress raises urgent questions around user consent and the ability to remove specific individuals from a model's output space. We address this by introducing SUGAR, a framework for scalable generative unlearning that enables the removal of many identities (simultaneously or sequentially) without retraining the entire model. Rather than projecting unwanted identities to unrealistic outputs or relying on static template faces, SUGAR learns a personalized surrogate latent for each identity, diverting reconstructions to visually coherent alternatives while preserving the model's quality and diversity. We further introduce a continual utility preservation objective that guards against degradation as more identities are forgotten. SUGAR achieves state-of-the-art performance in removing up to 200 identities, while delivering up to a 700% improvement in retention utility compared to existing baselines. Our code is publicly available at https://github.com/judydnguyen/SUGAR-Generative-Unlearn.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。