Mitigating Longitudinal Performance Degradation in Child Face Recognition Using Synthetic Data

TL;DR

This paper explores using synthetic face data generated by StyleGAN2 ADA to improve the longitudinal stability of child face recognition models, significantly reducing verification errors over time.

Contribution

It demonstrates that combining authentic and synthetic training data enhances child face recognition robustness against facial growth-related template drift.

Findings

Synthetic augmentation reduces error rates over 6 to 36 months

Fine-tuning with synthetic data improves temporal robustness

Synthetic data helps mitigate identity leakage and artifacts

Abstract

Longitudinal face recognition in children remains challenging due to rapid and nonlinear facial growth, which causes template drift and increasing verification errors over time. This work investigates whether synthetic face data can act as a longitudinal stabilizer by improving temporal robustness of child face recognition models. Using an identity disjoint protocol on the Young Face Aging (YFA) dataset, we evaluate three settings: (i) pretrained MagFace embeddings without dataset specific fine-tuning, (ii) MagFace fine-tuned using authentic training faces only, and (iii) MagFace fine-tuned using a combination of authentic and synthetically generated training faces. Synthetic data is generated using StyleGAN2 ADA and incorporated exclusively within the training identities; a post generation filtering step is applied to mitigate identity leakage and remove artifact affected samples. Experimental results across enrollment verification gaps from 6 to 36 months show that synthetic-augmented fine tuning substantially reduces error rates relative to both the pretrained baseline and real only fine tuning. These findings provide a risk aware assessment of synthetic augmentation for improving identity persistence in pediatric face recognition.