Abstract
The use of synthetic data as an alternative to authentic datasets in face recognition (FR) development has gained significant attention, addressing privacy, ethical, and practical concerns associated with collecting and using authentic data. Recent state-of-the-art approaches have proposed identity-conditioned diffusion models to generate identity-consistent face images, facilitating their use in training FR models. However, these methods often lack explicit sampling mechanisms to enforce inter-class separability, leading to identity overlap in the generated data and, consequently, suboptimal FR performance. In this work, we introduce NegFaceDiff, a novel sampling method that incorporates negative conditions into the identity-conditioned diffusion process. NegFaceDiff enhances identity separation by leveraging negative conditions that explicitly guide the model away from unwanted features while preserving intra-class consistency. Extensive experiments demonstrate that NegFaceDiff significantly improves the identity consistency and separability of data generated by identity-conditioned diffusion models. Specifically, identity separability, measured by the Fisher Discriminant Ratio (FDR), increases from 2.427 to 5.687. These improvements are reflected in FR systems trained on the NegFaceDiff dataset, which outperform models trained on data generated without negative conditions across multiple benchmarks.
Abstract (translated)
在人脸识别（FR）开发中，合成数据作为一种替代真实数据集的方法已经获得了广泛关注，这解决了收集和使用真实数据时出现的隐私、伦理以及实际操作方面的担忧。最近的研究前沿提出了身份条件扩散模型来生成一致的身份人脸图像，从而促进其用于训练人脸识别模型。然而，这些方法往往缺少明确的采样机制以强制执行跨类别的可分离性，导致生成的数据中存在身份重叠问题，并最终影响了人脸识别系统的性能。 在本研究中，我们引入了一种名为NegFaceDiff的新颖采样方法，该方法将负条件纳入到身份条件扩散过程中。通过使用负面条件明确引导模型远离不希望出现的特征同时保持同类内的一致性，NegFaceDiff增强了不同身份之间的分离度。广泛的实验表明，与没有负面条件的数据相比，使用NegFaceDiff生成的身份条件扩散数据在身份一致性和可分离性方面有了显著改善。具体来说，通过费舍尔判别比（FDR）测量的识别区分度从2.427提升到了5.687。 这些改进反映在基于NegFaceDiff数据集训练的人脸识别系统中，在多个基准测试上，此类模型的表现优于未使用负面条件生成的数据训练的模型。
URL
https://arxiv.org/abs/2508.09661
