analiza aiqbench
This paper presents SLOT-Align, a novel framework for One-Shot Federated Learning that resolves feature misalignment caused by simultaneous domain and label shifts using optimal transport. By leveraging frozen encoders and closed-form geodesic maps, the method offers an efficient way to harmonize heterogeneous client data without iterative retraining or heavy server-side computation. The approach demonstrates consistent gains in accuracy and robustness across multiple benchmarks, addressing a critical bottleneck in extreme communication-constrained AI deployment scenarios.
Distribution Alignment for One-Shot Federated Learning via Optimal Transport
arXiv:2606.16655v1 Announce Type: new Abstract: One-Shot Federated Learning (OSFL) addresses extreme communication regimes in which clients interact with the server only once, amplifying the impact of heterogeneous client data distributions. In particular, the interaction of domain shift and label shift across clients induces misaligned feature representations that cannot be corrected through iterative optimization. Existing OSFL methods rely on distillation, server-side generation or ensemble-based aggregation, but assume aligned representations or address domain and label shift separately. We introduce SLOT-Align (Single-round, Learning-free Optimal Transport Alignment), a geometry-aware feature harmonization framework for OSFL. SLOT-Align uses a shared frozen encoder to extract compact feature statistics, constructs a global reference via Bures-Wasserstein barycenters, and aligns local representations using closed-form geodesic optimal transport maps. The method is computationally efficient and can be combined with existing OSFL pipelines relying on frozen encoders without modifying their training procedures. Extensive experiments across multiple benchmarks, pretrained backbones, and OSFL methods show that SLOT-Align consistently improves accuracy and robustness under joint domain and label shift.
Abstract:One-Shot Federated Learning (OSFL) addresses extreme communication regimes in which clients interact with the server only once, amplifying the impact of heterogeneous client data distributions. In particular, the interaction of domain shift and label shift across clients induces misaligned feature representations that cannot be corrected through iterative optimization. Existing OSFL methods rely on distillation, server-side generation or ensemble-based aggregation, but assume aligned representations or address domain and label shift separately. We introduce SLOT-Align (Single-round, Learning-free Optimal Transport Alignment), a geometry-aware feature harmonization framework for OSFL. SLOT-Align uses a shared frozen encoder to extract compact feature statistics, constructs a global reference via Bures-Wasserstein barycenters, and aligns local representations using closed-form geodesic optimal transport maps. The method is computationally efficient and can be combined with existing OSFL pipelines relying on frozen encoders without modifying their training procedures. Extensive experiments across multiple benchmarks, pretrained backbones, and OSFL methods show that SLOT-Align consistently improves accuracy and robustness under joint domain and label shift.
Machine Learning (cs.LG)
arXiv:2606.16655 [cs.LG]
(or arXiv:2606.16655v1 [cs.LG] for this version)
https://doi.org/10.48550/arXiv.2606.16655 arXiv-issued DOI via DataCite (pending registration)
Text integral sub licența deschisă a publicației · 177 cuvinte
Subiecte conexe
News
Quantum Computing Report · 25 iul. 2026
News
Quantum Computing Report · 7 iun. 2026
Digital States 2026: AI Moves From Experiment to Infrastructure
govtech.com · 24 sept. 2026
FedeRage: Provably Convergent Agnostic Federated Learning under General Client Drift
arXiv · 21 sept. 2026