通过方向对齐发现面部识别嵌入中的交叉偏见
Discovering Intersectional Bias via Directional Alignment in Face Recognition Embeddings
AI总结：
本文提出LatentAlign算法，通过方向对齐发现面部识别嵌入中的语义连贯子群体，揭示隐藏的交叉偏见。
摘要：
现代面部识别模型将身份嵌入到单位超球面上，其中身份变化形成紧密簇群。相反，共享语义属性通常可以有效近似为潜在空间中的线性方向。现有偏见评估方法依赖预定义属性标签、合成反事实或基于接近度的聚类，均无法捕捉沿潜在方向出现的交叉子群体。我们引入LatentAlign，一种无属性算法，通过迭代对齐嵌入沿主导潜在方向，发现语义连贯且可解释的子群体。与基于距离的聚类不同，LatentAlign利用超球面嵌入的几何特性，隔离跨身份共享的方向结构，允许发现可解释的属性。在四个最先进的识别骨干网络（ArcFace、CosFace、ElasticFace、PartialFC）和两个基准（RFW、CelebA）上，LatentAlign始终比k-means、球面k-means、最近邻搜索和DBSCAN产生更语义连贯的组。关键的是，发现的子群体暴露了严重的交叉脆弱性，假匹配率高达显式标注组的4倍。我们的结果表明，通过将语义属性视为方向特征而非空间簇，可以有效隔离交叉子群体并揭示标准审计所遗漏的隐藏偏见。
英文摘要：
Modern face recognition models embed identities on a unit hypersphere, where identity variation forms tight clusters. Conversely, shared semantic attributes can often be effectively approximated as linear directions in the latent space. Existing bias evaluation methods rely on predefined attribute labels, synthetic counterfactuals, or proximity-based clustering, all of which fail to capture intersectional subpopulations that emerge along latent directions. We introduce LatentAlign, an attribute-free algorithm that discovers semantically coherent and interpretable subpopulations by iteratively aligning embeddings along dominant latent directions. Unlike distance-based clustering, LatentAlign exploits the geometry of hyperspherical embeddings to isolate directional structures shared across identities, allowing for the interpretable discovery of attributes. Across four state-of-the-art recognition backbones (ArcFace, CosFace, ElasticFace, PartialFC) and two benchmarks (RFW, CelebA), LatentAlign consistently yields more semantically coherent groups than $k$-means, spherical $k$-means, nearest-neighbor search, and DBSCAN. Crucially, the discovered subpopulations expose severe intersectional vulnerabilities, with False Match Rates up to 4x higher than groups defined by explicit annotations. Our results show that by treating semantic attributes as directional features rather than spatial clusters, we can effectively isolate intersectional subpopulations and expose hidden biases that standard audits miss.
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。
