GANDiff FR: Hybrid GAN Diffusion Synthesis for Causal Bias Attribution in Face Recognition

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Reaj, Md Asgor Hossain, Gupta, Rajan Das, Rahat, Md Yeasin, Fahad, Nafiz, Hasan, Md Jawadul, Liew, Tze Hui
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866915619678453760
author Reaj, Md Asgor Hossain
Gupta, Rajan Das
Rahat, Md Yeasin
Fahad, Nafiz
Hasan, Md Jawadul
Liew, Tze Hui
author_facet Reaj, Md Asgor Hossain
Gupta, Rajan Das
Rahat, Md Yeasin
Fahad, Nafiz
Hasan, Md Jawadul
Liew, Tze Hui
contents 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.
format Preprint
id arxiv_https___arxiv_org_abs_2508_11334
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle GANDiff FR: Hybrid GAN Diffusion Synthesis for Causal Bias Attribution in Face Recognition
Reaj, Md Asgor Hossain
Gupta, Rajan Das
Rahat, Md Yeasin
Fahad, Nafiz
Hasan, Md Jawadul
Liew, Tze Hui
Computer Vision and Pattern Recognition
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.
title GANDiff FR: Hybrid GAN Diffusion Synthesis for Causal Bias Attribution in Face Recognition
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2508.11334