联邦训练会降低可解释性吗？Grad-CAM 在集中式与联邦深度伪造检测中的忠实性对比分析
Does Federated Training Degrade Explainability? A Comparative Analysis of Grad-CAM Faithfulness in Centralized vs. Federated Deepfake Detection
AI 导读
一项 arXiv 研究系统评估了联邦深度伪造检测中 Grad-CAM 解释的忠实性，在 FaceForensics++ 数据集上用 EfficientNet-B4 检测器跨 5 个联邦客户端、不同数据异构程度下进行对比。
正文 · 原文
Abstract:Automated detection has become the essential safeguard against realistic synthetic media, but deploying it at scale requires training on sensitive data from multiple organizations, putting privacy and the transparency demanded by forensic review in direct conflict. Federated learning resolves the privacy side by sharing only model updates rather than raw data, but whether this comes at the cost of trustworthy gradient-based explanations, which analysts rely on to decide whether or not to trust a flagged clip, has not been studied. We present the systematic study of Grad-CAM explanation faithfulness under federated deepfake detection, evaluating an EfficientNet-B4 detector on FaceForensics++ across five federated clients under varying degrees of data heterogeneity. We find that the relationship between federation and explanation quality is non-monotonic: at near-IID (Independent and Identically Distributed Data) and moderate heterogeneity, federated models match or exceed the faithfulness of a centralized baseline, but at extreme heterogeneity, explanations break down entirely, becoming spatially incoherent and, on average, causally misleading. This degradation is not a general cost of federated training but a threshold effect tied to client heterogeneity, and we offer practical guidance for identifying when privacy-preserving deepfake detection can be deployed safely and when its explanations can no longer be trusted. Code and data partitions are available for reproducibility.
| Comments: | 10 Pages, 3 figures, Muhammad Uzair Shahid and Muhammad Zabil Mehboob contributed equally to this work |
| Subjects: | Artificial Intelligence (cs.AI); Cryptography and Security (cs.CR); Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG) |
| Cite as: | arXiv:2610.39483 [cs.AI] |
| (or arXiv:2610.39483v1 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2610.39483 arXiv-issued DOI via DataCite |
Submission history
From: Muhammad Zabil Mehboob [view email]
[v1]
Thu, 29 Oct 2026 06:49:14 UTC (516 KB)
来源：arXiv cs.AI · arxiv.org