通过模拟变形使任意事物不可分割
Unsegment Anything by Simulating Deformation
AI总结：
针对基础分割模型可能被滥用的问题，提出通过可微变形破坏图像编码器特征的 UAD 攻击，实现跨架构、跨提示接口的高可迁移性提示无关攻击。
摘要：
基础分割模型虽然功能强大，却带来了显著风险：它们使用户只需点击一下，就能毫不费力地从任何数字内容中提取任意对象，这可能导致版权侵犯或恶意滥用。为缓解这一风险，我们提出了一项名为“Anything Unsegmentable”的新任务，以赋予任何图像“不被分割的权利”。该任务的宏大目标是针对所有基于提示的分割模型实现高可迁移性对抗攻击，而不受模型参数化方式和提示的影响。我们强调了提示特定对抗噪声的不可迁移性和异质性。我们的方法聚焦于破坏图像编码器特征，以实现与提示无关的攻击。有趣的是，与无目标攻击相比，有目标特征攻击表现出更好的可迁移性，这表明最优更新方向与图像流形相一致。基于这些观察，我们设计了一种名为 Unsegment Anything by Simulating Deformation（UAD）的新型攻击。我们的攻击优化一个可微变形函数以生成目标变形图像，该图像会改变结构信息，同时保持对抗样本可实现的特征距离。大量实验验证了我们方法的有效性，该方法攻击了多种具有不同架构和提示接口的可提示分割模型。我们在 https://github.com/jiahaolu97/anything-unsegmentable 发布代码。
英文摘要：
Foundation segmentation models, while powerful, pose a significant risk: they enable users to effortlessly extract any objects from any digital content with a single click, potentially leading to copyright infringement or malicious misuse. To mitigate this risk, we introduce a new task "Anything Unsegmentable" to grant any image "the right to be unsegmented". The ambitious pursuit of the task is to achieve highly transferable adversarial attacks against all prompt-based segmentation models, regardless of model parameterizations and prompts. We highlight the non-transferable and heterogeneous nature of prompt-specific adversarial noises. Our approach focuses on disrupting image encoder features to achieve prompt-agnostic attacks. Intriguingly, targeted feature attacks exhibit better transferability compared to untargeted ones, suggesting the optimal update direction aligns with the image manifold. Based on the observations, we design a novel attack named Unsegment Anything by Simulating Deformation (UAD). Our attack optimizes a differentiable deformation function to create a target deformed image, which alters structural information while preserving achievable feature distance by adversarial example. Extensive experiments verify the effectiveness of our approach, compromising a variety of promptable segmentation models with different architectures and prompt interfaces. We release the code at https://github.com/jiahaolu97/anything-unsegmentable.
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。
