aiqbench analysis
This paper advances federated learning by introducing geometry-aware aggregation strategies that preserve the intrinsic structure of symmetric positive definite matrices, addressing a critical limitation in standard Euclidean averaging methods. By projecting means onto the Stiefel manifold or using tangent-space approximations, the proposed frameworks achieve superior performance and robustness on EEG motor imagery benchmarks while reducing communication overhead. The work represents a significant methodological step for deploying machine learning models in distributed settings where data privacy and geometric constraints are paramount.
FedSPDnet: Geometry-Aware Federated Deep Learning with SPDnet
arXiv:2604.22494v2 Announce Type: replace-cross Abstract: We introduce two federated learning frameworks for the classical SPDnet model operating on symmetric positive definite (SPD) matrices with Stiefel-constrained parameters. Unlike standard Euclidean averaging, which violates orthogonality, our approach preserves geometric structure through two efficient aggregation strategies: ProjAvg, projecting arithmetic means onto the Stiefel manifold, and RLAvg, approximating tangent-space averaging via retractions and liftings. Both methods are computationally efficient, independent of the optimizer, and enable scalable federated learning for signal processing applications whose features are SPD matrices. Simulations on EEG motor imagery benchmarks show that FedSPDnet outperforms federated EEGnet in F1 score and robustness to federation and partial participation, while using fewer parameters per communication round.
Abstract:We introduce two federated learning frameworks for the classical SPDnet model operating on symmetric positive definite (SPD) matrices with Stiefel-constrained parameters. Unlike standard Euclidean averaging, which violates orthogonality, our approach preserves geometric structure through two efficient aggregation strategies: ProjAvg, projecting arithmetic means onto the Stiefel manifold, and RLAvg, approximating tangent-space averaging via retractions and liftings. Both methods are computationally efficient, independent of the optimizer, and enable scalable federated learning for signal processing applications whose features are SPD matrices. Simulations on EEG motor imagery benchmarks show that FedSPDnet outperforms federated EEGnet in F1 score and robustness to federation and partial participation, while using fewer parameters per communication round.
Machine Learning (stat.ML); Machine Learning (cs.LG)
arXiv:2604.22494 [stat.ML]
(or arXiv:2604.22494v2 [stat.ML] for this version)
https://doi.org/10.48550/arXiv.2604.22494 Focus to learn more arXiv-issued DOI via DataCite
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 202 words
Related coverage
News
Quantum Computing Report · Jul 25, 2026
QSplitFL: Capability Aware Deep Q-Learning for Optimal Split Point Selection in Split Federated Learning
arXiv · Jun 10, 2026
Geometric origin of adversarial vulnerability in deep learning
arXiv · Jul 21, 2026
JGRA: Jacobian Geometry Robustness Assessment in NISQ Noise-Aware Quantum Neural Networks
arXiv · Jun 12, 2026