LDP-Slicing：通过随机位平面切片实现图像的局部差分隐私
LDP-Slicing: Local Differential Privacy for Images via Randomized Bit-Plane Slicing
把像素值拆成二值位平面后再施加 LDP 机制，为视觉任务提供了一种免训练、可直接插在数据采集端的隐私保护预处理；做隐私敏感场景（如人脸识别数据外包、联邦式图像采集）时，可复用该位级扰动+预算分配思路替代直接在像素上加噪。
原文依据 decompose pixel values into a sequence of binary bit-planes
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论局部差分隐私（LDP）通过在数据源头保证隐私，是隐私保护机器学习的黄金标准信任模型。然而，由于像素空间的高维性，其在图像数据上的应用长期被认为不切实际。经典的 LDP 机制是为低维数据设计的，应用于高维像素空间时会导致严重的效用退化。本文表明，这种效用损失并非 LDP 固有的，而是源于将其施加于不恰当的数据表示。我们提出 LDP-Slicing，一个轻量、免训练的框架，用以解决这一领域失配问题。我们的关键洞见是将像素值分解为一系列二值位平面。这一变换使我们能够直接将 LDP 机制作用于位级表示。为进一步增强隐私并保持效用，我们引入了一个感知混淆模块以缓解人眼可感知的信息泄露，以及一种基于优化的隐私预算分配策略。该流程在满足严格的像素级 ε-LDP 的同时，生成的图像对下游任务仍保持高效用。在人脸识别与图像分类上的大量实验表明，在相当的隐私预算下，LDP-Slicing 优于现有的 DP/LDP 基线，且计算开销可忽略不计。
英文原文摘要 1382 字符
Local Differential Privacy (LDP) is the gold standard trust model for privacy-preserving machine learning by guaranteeing privacy at the data source. However, its application to image data has long been considered impractical due to the high dimensionality of pixel space. Canonical LDP mechanisms are designed for low-dimensional data, resulting in severe utility degradation when applied to high-dimensional pixel spaces. This paper demonstrates that this utility loss is not inherent to LDP, but from its application to an inappropriate data representation. We introduce LDP-Slicing, a lightweight, training-free framework that resolves this domain mismatch. Our key insight is to decompose pixel values into a sequence of binary bit-planes. This transformation allows us to apply the LDP mechanism directly to the bit-level representation. To further strengthen privacy and preserve utility, we integrate a perceptual obfuscation module that mitigates human-perceivable leakage and an optimization-based privacy budget allocation strategy. This pipeline satisfies rigorous pixel-level \varepsilon-LDP while producing images that retain high utility for downstream tasks. Extensive experiments on face recognition and image classification demonstrate that LDP-Slicing outperforms existing DP/LDP baselines under comparable privacy budgets, with negligible computational overhead.