首个可精确控制人脸属性的合成框架，用于公平性审计与偏差归因。
GANDiff FR: Hybrid GAN Diffusion Synthesis for Causal Bias Attribution in Face Recognition
- 融合风格化GAN与扩散模型，实现姿态、光照、表情的精细调控。
- 合成1万张均衡人脸，发现光照贡献42%残余偏差，AdaFace降低60%组间误判率。
- 支持合规审计，生成数据可跨数据集迁移，适合研究公平性的团队使用。
我们提出GANDiff FR，首个能精确控制人口统计与环境因素的合成框架，用于测量、解释和减少人脸识别中的偏见，并具备可复现的严谨性。该框架结合StyleGAN3的身份保持生成与基于扩散的属性控制，可在其他条件不变的情况下，对姿态（约30度）、光照（四个方向）和表情（五级）进行细粒度操控。我们合成10,000张跨五个群体的均衡人脸，通过自动化检测（98.2%）和人工评估（89%）验证其真实性，以隔离并量化偏差驱动因素。在匹配操作点下对比ArcFace、CosFace和AdaFace，结果显示AdaFace将组间真阳性率差异降低60%（2.5% vs. 6.3%），光照贡献了42%的残余偏差。在RFW、BUPT和CASIA WebFace上的跨数据集评估表明合成到真实迁移效果良好（r=0.85）。尽管相较纯GAN增加约20%计算开销，但生成的属性条件样本数量提升三倍，确立了符合欧盟人工智能法案（EU AI Act）的可复现公平性审计标准。代码与数据已公开，支持透明、可扩展的偏见评估。
原文摘要 · Abstract (English)
We introduce GANDiff FR, the first synthetic framework that precisely controls demographic and environmental factors to measure, explain, and reduce bias with reproducible rigor. GANDiff FR unifies StyleGAN3-based identity-preserving generation with diffusion-based attribute control, enabling fine-grained manipulation of pose around 30 degrees, illumination (four directions), and expression (five levels) under ceteris paribus conditions. We synthesize 10,000 demographically balanced faces across five cohorts validated for realism via automated detection (98.2%) and human review (89%) to isolate and quantify bias drivers. Benchmarking ArcFace, CosFace, and AdaFace under matched operating points shows AdaFace reduces inter-group TPR disparity by 60% (2.5% vs. 6.3%), with illumination accounting for 42% of residual bias. Cross-dataset evaluation on RFW, BUPT, and CASIA WebFace confirms strong synthetic-to-real transfer (r 0.85). Despite around 20% computational overhead relative to pure GANs, GANDiff FR yields three times more attribute-conditioned variants, establishing a reproducible, regulation-aligned (EU AI Act) standard for fairness auditing. Code and data are released to support transparent, scalable bias evaluation.
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。