ID$^3$: Identity-Preserving-yet-Diversified Diffusion Models for Synthetic Face Recognition

TL;DR

The paper introduces ID$^3$, a diffusion-based model for synthetic face recognition that generates diverse, identity-preserving face datasets to improve privacy and model training, with theoretical and experimental validation.

Contribution

We propose a novel diffusion model, ID$^3$, that maintains identity while enhancing diversity in synthetic face data, supported by a new loss function and sampling algorithm.

Findings

Outperforms existing methods on five benchmarks

Generates highly diverse and identity-consistent faces

Theoretically grounded in maximizing a lower bound of likelihood

Abstract

Synthetic face recognition (SFR) aims to generate synthetic face datasets that mimic the distribution of real face data, which allows for training face recognition models in a privacy-preserving manner. Despite the remarkable potential of diffusion models in image generation, current diffusion-based SFR models struggle with generalization to real-world faces. To address this limitation, we outline three key objectives for SFR: (1) promoting diversity across identities (inter-class diversity), (2) ensuring diversity within each identity by injecting various facial attributes (intra-class diversity), and (3) maintaining identity consistency within each identity group (intra-class identity preservation). Inspired by these goals, we introduce a diffusion-fueled SFR model termed $\text{ID}^3$. $\text{ID}^3$ employs an ID-preserving loss to generate diverse yet identity-consistent facial appearances. Theoretically, we show that minimizing this loss is equivalent to maximizing the lower bound of an adjusted conditional log-likelihood over ID-preserving data. This equivalence motivates an ID-preserving sampling algorithm, which operates over an adjusted gradient vector field, enabling the generation of fake face recognition datasets that approximate the distribution of real-world faces. Extensive experiments across five challenging benchmarks validate the advantages of $\text{ID}^3$.