mlsanityData to 7 Oct 2026

HEART-PFL: Stable Personalized Federated Learning under Heterogeneity with Hierarchical Directional Alignment and Adversarial Knowledge Transfer

arXiv 2603.24209 ↗First posted 25 Mar 2026cs.CV cs.LGcited by 0

Not on the map. Why?

Abstract (arXiv)

Personalized Federated Learning (PFL) aims to deliver effective client-specific models under heterogeneous distributions, yet existing methods suffer from shallow prototype alignment and brittle server-side distillation. We propose HEART-PFL, a dual-sided framework that (i) performs depth-aware Hierarchical Directional Alignment (HDA) using cosine similarity in the early stage and MSE matching in the deep stage to preserve client specificity, and (ii) stabilizes global updates through Adversarial Knowledge Transfer (AKT) with symmetric KL distillation on clean and adversarial proxy data. Using lightweight adapters with only 1.46M trainable parameters, HEART-PFL achieves state-of-the-art personalized accuracy on CIFAR-100, Flowers-102, and Caltech-101 (63.42%, 84.23%, and 95.67%, respectively) under Dirichlet non-IID partitions, and remains robust to out-of-domain proxy data. Ablation studies further confirm that HDA and AKT provide complementary gains in alignment, robustness, and optimization stability, offering insights into how the two components mutually reinforce effective personalization. Overall, these results demonstrate that HEART-PFL simultaneously enhances personalization and global stability, highlighting its potential as a strong and scalable solution for PFL(code available at https://github.com/danny0628/HEART-PFL).

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Builds on 34 of its 44 references that are in the map. First posted on arXiv 25 Mar 2026.

Builds on

34 of 44 references.

References · 34

  1. 2020A Simple Framework for Contrastive Learning of Visual Representationscited by 11,979
  2. 2015Distilling the Knowledge in a Neural Networkcited by 10,076
  3. 2014ImageNet Large Scale Visual Recognition Challengecited by 8,441
  4. 2016Communication-Efficient Learning of Deep Networks from Decentralized Datacited by 7,592
  5. 2017Towards Deep Learning Models Resistant to Adversarial Attackscited by 6,326
  6. 2018Federated Optimization in Heterogeneous Networkscited by 2,971
  7. 2019On the Convergence of FedAvg on Non-IID Datacited by 1,220
  8. 2021Model-Contrastive Federated Learningcited by 762
  9. 2020Ensemble Distillation for Robust Model Fusion in Federated Learningcited by 678
  10. 2020Ditto: Fair and Robust Federated Learning Through Personalizationcited by 665
  11. 2020Personalized Federated Learning with Moreau Envelopescited by 612
  12. 2021Exploiting Shared Representations for Personalized Federated Learningcited by 537
22 more
  1. 2019Federated Learning with Personalization Layerscited by 533
  2. 2019FedMD: Heterogenous Federated Learning via Model Distillationcited by 454
  3. 2021FedProto: Federated Prototype Learning across Heterogeneous Clientscited by 349
  4. 2020Think Locally, Act Globally: Federated Learning with Local and Global Representationscited by 322
  5. 2020Three Approaches for Personalization with Applications to Federated Learningcited by 321
  6. 2020Federated Learning of a Mixture of Global and Local Modelscited by 208
  7. 2022FedDC: Federated Learning with Non-IID Data via Local Drift Decoupling and Correctioncited by 181
  8. 2021FedBABU: Towards Enhanced Representation for Federated Image Classificationcited by 180
  9. 2020Lower Bounds and Optimal Algorithms for Personalized Federated Learningcited by 122
  10. 2022FedALA: Adaptive Local Aggregation for Personalized Federated Learningcited by 121
  11. 2022Federated Learning with Partial Model Personalizationcited by 114
  12. 2021Parameterized Knowledge Transfer for Personalized Federated Learningcited by 108
  13. 2022FedDM: Iterative Distribution Matching for Communication-Efficient Federated Learningcited by 65
  14. 2024Rethinking Kullback-Leibler Divergence in Knowledge Distillation for Large Language Modelscited by 60
  15. 2023FedCP: Separating Feature Information for Personalized Federated Learning via Conditional Policycited by 36
  16. 2024FedSelect: Personalized Federated Learning with Customized Selection of Parameters for Fine-Tuningcited by 27
  17. 2023Understanding the Feature Norm for Out-of-Distribution Detectioncited by 25
  18. 2022Self-Knowledge Distillation via Dropoutcited by 18
  19. 2023GPFL: Simultaneously Learning Global and Personalized Feature Information for Personalized Federated Learningcited by 18
  20. 2023PerAda: Parameter-Efficient Federated Learning Personalization with Generalization Guaranteescited by 15
  21. 2025Balance Divergence for Knowledge Distillationcited by 2
  22. 2025FedORGP: Guiding Heterogeneous Federated Learning with Orthogonality Regularization on Global Prototypescited by 1

Built on by

Not cited yet.