FedUNet: A Lightweight Additive U-Net Module for Federated Learning with Heterogeneous Models
Fuente:
arXiv
Salvato in:
| Autori principali: | , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2025
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866908493021184000 |
|---|---|
| author | Seo, Beomseok Lee, Kichang Park, JaeYeon |
| author_facet | Seo, Beomseok Lee, Kichang Park, JaeYeon |
| contents | Federated learning (FL) enables decentralized model training without sharing local data. However, most existing methods assume identical model architectures across clients, limiting their applicability in heterogeneous real-world environments. To address this, we propose FedUNet, a lightweight and architecture-agnostic FL framework that attaches a U-Net-inspired additive module to each client's backbone. By sharing only the compact bottleneck of the U-Net, FedUNet enables efficient knowledge transfer without structural alignment. The encoder-decoder design and skip connections in the U-Net help capture both low-level and high-level features, facilitating the extraction of clientinvariant representations. This enables cooperative learning between the backbone and the additive module with minimal communication cost. Experiment with VGG variants shows that FedUNet achieves 93.11% accuracy and 92.68% in compact form (i.e., a lightweight version of FedUNet) with only 0.89 MB low communication overhead. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2508_12740 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | FedUNet: A Lightweight Additive U-Net Module for Federated Learning with Heterogeneous Models Seo, Beomseok Lee, Kichang Park, JaeYeon Machine Learning Artificial Intelligence 68T01 (Primary), 68T07 (Secondary) I.2 Federated learning (FL) enables decentralized model training without sharing local data. However, most existing methods assume identical model architectures across clients, limiting their applicability in heterogeneous real-world environments. To address this, we propose FedUNet, a lightweight and architecture-agnostic FL framework that attaches a U-Net-inspired additive module to each client's backbone. By sharing only the compact bottleneck of the U-Net, FedUNet enables efficient knowledge transfer without structural alignment. The encoder-decoder design and skip connections in the U-Net help capture both low-level and high-level features, facilitating the extraction of clientinvariant representations. This enables cooperative learning between the backbone and the additive module with minimal communication cost. Experiment with VGG variants shows that FedUNet achieves 93.11% accuracy and 92.68% in compact form (i.e., a lightweight version of FedUNet) with only 0.89 MB low communication overhead. |
| title | FedUNet: A Lightweight Additive U-Net Module for Federated Learning with Heterogeneous Models |
| topic | Machine Learning Artificial Intelligence 68T01 (Primary), 68T07 (Secondary) I.2 |
| url | https://arxiv.org/abs/2508.12740 |