发表机构
ELLIS Institute Finland; Center for Machine Vision and Signal Analysis (CMVS), University of Oulu（芬兰ELLIS研究所; 奥卢大学机器视觉与信号分析中心（CMVS））
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
本研究提出统一的人脸去识别化评估框架，包括平衡基准UtilFace和分层度量HiFD，以解决现有评估碎片化问题，并揭示不同方法间的权衡与失败模式。
AI 中文摘要
人脸去识别化（FDeID）已成为一项关键的隐私保护技术，然而其评估在根本上仍是碎片化的。现有协议依赖于不一致的度量标准、异构数据集以及部分标注覆盖，因此针对不同效用维度（如地标保留与表情保留）的方法，在不同基准上采用不同度量进行报告，导致跨方法比较不可行。我们从数据和度量两个角度重新审视FDeID评估。在数据方面，我们引入了UtilFace，一个经过精心整理、人口统计平衡且具有高身份多样性的基准，通过身份感知清洗、分辨率增强和分层过滤，从四个大规模人脸数据集中构建而成。在度量方面，我们提出了HiFD，一种分层人脸去识别化度量，将身份抑制、多级效用保留和图像质量统一在单一的一致性范式下：每个组件均基于预训练估计器对原始人脸及其去识别化对应物输出的计算，直接量化身份被抑制的程度以及下游可感知效用存留的程度。HiFD将人脸信号组织为三级效用层次，涵盖宏观线索（L1）、微观线索（L2）和不可感知线索（L3），并通过加权调和平均将五个结果组件聚合成单一可解释分数，支持可配置的应用特定配置文件。利用这一统一协议，我们开展了一项涵盖对抗性、基于GAN和基于扩散方法的全面比较研究，揭示了现有协议下不可见的权衡和失败模式。我们发布了该基准和评估工具包，以促进隐私保护人脸分析中系统化和可重复的研究。
英文摘要
Face de-identification (FDeID) has emerged as a critical privacy-preserving technology, yet its evaluation remains fundamentally fragmented. Existing protocols rely on inconsistent metrics, heterogeneous datasets, and partial annotation coverage, so methods targeting different utility dimensions, such as landmark versus expression preservation, are reported on different benchmarks under different metrics, rendering cross-method comparison infeasible. We revisit FDeID evaluation from both the data and metric perspectives. On the data side, we introduce UtilFace, a curated, demographically balanced benchmark with high identity diversity, assembled from four large-scale face datasets through identity-aware cleaning, resolution enhancement, and stratified filtering. On the metric side, we propose HiFD, a Hierarchical Face De-identification metric that unifies identity suppression, multi-level utility preservation, and image quality under a single consistency-based paradigm: every component is computed from pretrained estimators' outputs on the original face and its de-identified counterpart, directly quantifying how much identity is suppressed and how much downstream-perceivable utility survives. HiFD organizes facial signals into a three-level utility hierarchy spanning macro cues (L1), micro cues (L2), and imperceptible cues (L3), and aggregates the five resulting components into a single interpretable score via weighted harmonic mean, with configurable application-specific profiles. Using this unified protocol, we conduct a comprehensive comparative study spanning adversarial, GAN-based, and diffusion-based methods, surfacing trade-offs and failure modes that remain invisible under existing protocols. We release the benchmark and evaluation toolkit to foster systematic and reproducible research in privacy-preserving human face analysis.
CommentsAccepted to NeurIPS 2026. Project Page: https://cv-ac.github.io/hifd/
