mlsanityData to 14 Oct 2026

Multi-metrics adaptively identifies backdoors in Federated learning

arXiv 2303.06601 ↗First posted 12 Mar 2023cs.CR cs.CV cs.LGcited by 23

Not on the map. Why?

Abstract (arXiv)

The decentralized and privacy-preserving nature of federated learning (FL) makes it vulnerable to backdoor attacks aiming to manipulate the behavior of the resulting model on specific adversary-chosen inputs. However, most existing defenses based on statistical differences take effect only against specific attacks, especially when the malicious gradients are similar to benign ones or the data are highly non-independent and identically distributed (non-IID). In this paper, we revisit the distance-based defense methods and discover that i) Euclidean distance becomes meaningless in high dimensions and ii) malicious gradients with diverse characteristics cannot be identified by a single metric. To this end, we present a simple yet effective defense strategy with multi-metrics and dynamic weighting to identify backdoors adaptively. Furthermore, our novel defense has no reliance on predefined assumptions over attack settings or data distributions and little impact on benign performance. To evaluate the effectiveness of our approach, we conduct comprehensive experiments on different datasets under various attack settings, where our method achieves the best defensive performance. For instance, we achieve the lowest backdoor accuracy of 3.06% under the difficult Edge-case PGD, showing significant superiority over previous defenses. The results also demonstrate that our method can be well-adapted to a wide range of non-IID degrees without sacrificing the benign performance.

Key facts. Cited by 23 papers in the map (data through 14 Oct 2026). Builds on 32 of its 66 references that are in the map. First posted on arXiv 12 Mar 2023.

Builds on

32 of 66 references.

References · 32

  1. 2020Language Models are Few-Shot Learnerscited by 32,419
  2. 2014Very Deep Convolutional Networks for Large-Scale Image Recognitioncited by 15,364
  3. 2016Communication-Efficient Learning of Deep Networks from Decentralized Datacited by 7,669
  4. 2022Hierarchical Text-Conditional Image Generation with CLIP Latentscited by 6,585
  5. 2021Switch Transformers: Scaling to Trillion Parameter Models with Simple and Efficient Sparsitycited by 3,073
  6. 2019Advances and Open Problems in Federated Learningcited by 3,046
  7. 2016Federated Learning: Strategies for Improving Communication Efficiencycited by 1,475
  8. 2022CoCa: Contrastive Captioners are Image-Text Foundation Modelscited by 1,274
  9. 2021Swin Transformer V2: Scaling Up Capacity and Resolutioncited by 1,228
  10. 2017BadNets: Identifying Vulnerabilities in the Machine Learning Model Supply Chaincited by 1,106
  11. 2017Targeted Backdoor Attacks on Deep Learning Systems Using Data Poisoningcited by 1,055
  12. 2018How To Backdoor Federated Learningcited by 765
20 more
  1. 2021Scaling Vision with Sparse Mixture of Expertscited by 747
  2. 2018Byzantine-Robust Distributed Learning: Towards Optimal Statistical Ratescited by 711
  3. 2020Invisible Backdoor Attack with Sample-Specific Triggerscited by 353
  4. 2018Analyzing Federated Learning through an Adversarial Lenscited by 351
  5. 2020FLTrust: Byzantine-robust Federated Learning via Trust Bootstrappingcited by 346
  6. 2018The Hidden Vulnerability of Distributed Learning in Byzantiumcited by 329
  7. 2019Robust Aggregation for Federated Learningcited by 317
  8. 2020Attack of the Tails: Yes, You Really Can Backdoor Federated Learningcited by 309
  9. 2019Can You Really Backdoor Federated Learning?cited by 255
  10. 2021CRFL: Certifiably Robust Federated Learning against Backdoor Attackscited by 91
  11. 2020Learning to Detect Malicious Clients for Robust Federated Learningcited by 88
  12. 2021OoD-Bench: Quantifying and Understanding Two Dimensions of Out-of-Distribution Generalizationcited by 86
  13. 2019Generative Models for Effective ML on Private, Decentralized Datasetscited by 85
  14. 2020Defending against Backdoors in Federated Learning with Robust Learning Ratecited by 82
  15. 2021Collaborative Unsupervised Visual Representation Learning from Decentralized Datacited by 69
  16. 2020BaFFLe: Backdoor detection via Feedback-based Federated Learningcited by 53
  17. 2019Attack-Resistant Federated Learning with Residual-based Reweightingcited by 40
  18. 2021INTERN: A New Learning Paradigm Towards General Visioncited by 33
  19. 2021More than Encoder: Introducing Transformer Decoder to Upsamplecited by 26
  20. 2020Fractional norms and quasinorms do not help to overcome the curse of dimensionalitycited by 13

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2023: 4; 2024: 3; 2025: 7; 2026: 9.

23 papers.

Sort:

By area:

  1. 2023Federated Learning for Generalization, Robustness, Fairness: A Survey and Benchmarkcited by 99
  2. 2023Diverse Cotraining Makes Strong Semi-Supervised Segmentorcited by 26
  3. 2023Defenses in Adversarial Machine Learning: A Surveycited by 14
  4. 2025Detecting Backdoor Attacks in Federated Learning via Direction Alignment Inspectioncited by 13
  5. 2024Identify Backdoored Model in Federated Learning via Individual Unlearningcited by 4
  6. 2025Bad-PFL: Exploring Backdoor Attacks against Personalized Federated Learningcited by 2
  7. 2023Towards Understanding Adversarial Transferability in Federated Learningcited by 1
  8. 2025Defending the Edge: Representative-Attention Defense against Backdoor Attacks in Federated Learningcited by 1
  9. 2025BackFed: An Efficient & Standardized Benchmark Suite for Backdoor Attacks in Federated Learningcited by 1
  10. 2025On the Out-of-Distribution Backdoor Attack for Federated Learningcited by 1
  11. 2026BackdoorIDS: Zero-shot Backdoor Detection for Pretrained Vision Encodercited by 1
  12. 2024FLAIN: Mitigating Backdoor Attacks in Federated Learning via Flipping Weight Updates of Low-Activation Input Neuronscited by 0
11 more
  1. 2024LOGSAFE: Logic-Guided Verification for Trustworthy Federated Time-Series Learningcited by 0
  2. 2025BadPromptFL: A Novel Backdoor Threat to Prompt-based Federated Learning in Multimodal Modelscited by 0
  3. 2025Hammer and Anvil: Toward a Theory of Backdoors in Federated Learningcited by 0
  4. 2026FAROS: Robust Federated Learning with Adaptive Scaling against Backdoor Attackscited by 0
  5. 2026CoRe-Fed: Bridging Collaborative and Representation Fairness via Federated Embedding Distillationcited by 0
  6. 2026TASER: Task-Aware Spectral Energy Refine for Backdoor Suppression in UAV Swarms Decentralized Federated Learningcited by 0
  7. 2026Defending Against Backdoor Attacks via Alignment Checking in Model-Contrastive Federated Learningcited by 0
  8. 2026FL-OA: A Byzantine-Robust Federated Learning Framework with Outsourced Auditing for Intelligent Devicescited by 0
  9. 2026Robust Reputation-Driven Crowdsourced Federated Learningcited by 0
  10. 2026BackDFL: A Unified Benchmark For Backdoor Attacks and Defenses In Decentralized Federated Learningcited by 0
  11. 2026Backdoors Leave Structural Traces: FedMAST for Backdoor Detection and Containment in Federated Learningcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2023Avoid Adversarial Adaption in Federated Learning by Multi-Metric Investigations
  2. 2023On the Vulnerability of Backdoor Defenses for Federated Learning
  3. 2022Identifying Backdoor Attacks in Federated Learning via Anomaly Detection
  4. 2025MARS: A Malignity-Aware Backdoor Defense in Federated Learning
  5. 2025FedBAP: Backdoor Defense via Benign Adversarial Perturbation in Federated Learning
  6. 2025Runtime Backdoor Detection for Federated Learning via Representational Dissimilarity Analysis
  7. 2025Enhancing the Effectiveness and Durability of Backdoor Attacks in Federated Learning through Maximizing Task Distinction
  8. 2024Client-Side Patching against Backdoor Attacks in Federated Learning