A Race Bias Free Face Aging Model for Reliable Kinship Verification

2025-09-18

AI Summary

To address the challenges of large age gaps in parent-child photographs and algorithmic bias in existing cross-racial facial aging models—which degrade kinship verification accuracy and fairness—this paper proposes RA-GAN, the first generative method incorporating a race-unbiased facial aging mechanism. RA-GAN integrates the RACEpSp encoder with a feature-mixing module to generate high-fidelity, identity-preserving same-age images, effectively mitigating performance degradation in multi-ethnic scenarios. On the KinFaceW-I/II benchmarks, RA-GAN achieves up to a 5.22% improvement in kinship verification accuracy. For subjects aged 60+, it attains a 9.1% higher cross-racial verification accuracy than CUSP-GAN, while also outperforming both SAM-GAN and CUSP-GAN in identity preservation. These results demonstrate RA-GAN’s superior generalizability, fairness, and robustness across diverse demographic groups.

Technology Category

• Computer Vision: Generative Adversarial Networks (GANs) for Vision
• Machine Learning: Deep Generative Models & Autoencoders
• Cognitive Modeling & Cognitive Systems: Affective Computing

Application Category

• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Social Networks and Social Media: Generative AI / large language models and their impact on social systems

Abstract

The age gap in kinship verification addresses the time difference between the photos of the parent and the child. Moreover, their same-age photos are often unavailable, and face aging models are racially biased, which impacts the likeness of photos. Therefore, we propose a face aging GAN model, RA-GAN, consisting of two new modules, RACEpSp and a feature mixer, to produce racially unbiased images. The unbiased synthesized photos are used in kinship verification to investigate the results of verifying same-age parent-child images. The experiments demonstrate that our RA-GAN outperforms SAM-GAN on an average of 13.14% across all age groups, and CUSP-GAN in the 60+ age group by 9.1% in terms of racial accuracy. Moreover, RA-GAN can preserve subjects' identities better than SAM-GAN and CUSP-GAN across all age groups. Additionally, we demonstrate that transforming parent and child images from the KinFaceW-I and KinFaceW-II datasets to the same age can enhance the verification accuracy across all age groups. The accuracy increases with our RA-GAN for the kinship relationships of father-son and father-daughter, mother-son, and mother-daughter, which are 5.22, 5.12, 1.63, and 0.41, respectively, on KinFaceW-I. Additionally, the accuracy for the relationships of father-daughter, father-son, and mother-son is 2.9, 0.39, and 1.6 on KinFaceW-II, respectively. The code is available at~href{https://github.com/bardiya2254kariminia/An-Age-Transformation-whitout-racial-bias-for-Kinship-verification}{Github}

Problem

• Addressing racial bias in face aging models
• Improving kinship verification with same-age photos
• Enhancing identity preservation across age groups

Innovation

• RA-GAN model with RACEpSp module
• Feature mixer for racially unbiased images
• Same-age photo synthesis for kinship verification