arXiv cs.AI· Ye Lu, Shen Wang, Zhaoyang Zhang, Yihan Yan, Li Liu, Runze Liu, Fanghui Sun·· 14 小时前AI 评分29
SFMI：用梯度引导流匹配反演人脸识别模型
Steering the Flow: Inverting Face Recognition Models via Gradient-Guided Flow Matching
AI 导读
研究者提出两阶段白盒模型反演方法 SFMI，将反演重构为轨迹引导任务：先预训练通用无条件 Flow Matching 模型编码人脸流形先验，再在采样中用渐进引导调度器（PGS）注入随时间变化的目标特定梯度。在 CelebA 身份不交叉评估下，对 ArcFace 目标模型取得 ACC 0.9248、FID 22.61、LPIPS 0.3874。
正文 · 原文
Abstract:Model Inversion Attacks (MIAs) aim to reconstruct representative training samples of target identities from face recognition models, exposing critical security vulnerabilities. Existing methods typically rely on indirect guidance or highly stochastic guidance, making it difficult to stably optimize generation trajectories toward target facial images. In this paper, we propose Steering Flow Model Inversion (SFMI), a novel two-stage white-box model inversion method that reformulates inversion as a trajectory-steering task. Specifically, Step I, Learning a Generic Flow Matching Prior, pre-trains a generic unconditional Flow Matching model to encode the manifold of human faces as a robust prior. Step II, Attacking with Progressive Guidance Scheduler (PGS), injects time-dependent target-specific gradients during sampling. By backpropagating through the target model to obtain gradients from intermediate generated states, PGS progressively injects adaptive guidance signals into the vector field. This process effectively steers the current generative flow from random noise toward the high-density regions of the target class. Under an identity-disjoint cross-evaluation setting using the CelebA dataset, SFMI achieves an ACC of 0.9248, an FID of 22.61, and an LPIPS of 0.3874 on the ArcFace target. Extensive experiments on multiple target models demonstrate that SFMI achieves competitive state-of-the-art performance in attack success and visual fidelity under the evaluated white-box protocol.
| Comments: | Accepted for publication in IEEE Internet of Things Journal. Updated to the author-accepted manuscript; 15 pages, 6 figures, including appendix |
| Subjects: | Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI); Cryptography and Security (cs.CR); Multimedia (cs.MM) |
| Cite as: | arXiv:2608.16791 [cs.CV] |
| (or arXiv:2608.16791v2 [cs.CV] for this version) | |
| https://doi.org/10.48550/arXiv.2608.16791 arXiv-issued DOI via DataCite |
|
| Related DOI: | https://doi.org/10.1109/JIOT.2026.3742815
DOI(s) linking to related resources |
Submission history
From: Ye Lu [view email]
[v1]
Mon, 17 Aug 2026 16:47:09 UTC (6,916 KB)
[v2]
Sun, 25 Oct 2026 09:43:06 UTC (8,733 KB)
来源：arXiv cs.AI · arxiv.org