Abstract
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.
Abstract (translated)
我们介绍了GANDiff FR，这是第一个能够精确控制人口统计学和环境因素的合成框架，用于测量、解释和减少偏差，并且具有可重复验证的严谨性。GANDiff FR结合了基于StyleGAN3的身份保持生成与基于扩散的方法属性控制，使得在其他条件不变的情况下，可以精细调整姿态（约30度范围内）、照明（四个方向）以及表情（五个级别）。我们合成了10,000张跨五类群体的人口统计学平衡面部图像，并通过自动检测（98.2%）和人工审核（89%）验证了其真实感，以便隔离并量化偏差驱动因素。在匹配的操作点下对ArcFace、CosFace和AdaFace进行基准测试显示，AdaFace将组间TPR差异减少了60%（2.5% vs 6.3%），其中照明占剩余偏见的42%。跨数据集评估证明了RFW、BUPT和CASIA WebFace中强大的合成到真实图像转移能力（相关系数r为0.85）。尽管相对于纯GANs，GANDiff FR在计算上增加了约20%的开销，但它生成的属性条件变体是纯GAN的三倍，并建立了符合欧盟人工智能法案等监管标准的可重复且公平审计的标准。代码和数据已发布以支持透明、可扩展的偏差评估。
URL
https://arxiv.org/abs/2508.11334
