Federated Learning for Computer Vision


Adaptive Differential Filters for Fast and Communication-Efficient Federated Learning
Daniel Becking,
Heiner Kirchhoffer,
Gerhard Tech,
Paul Haase,
Karsten Müller,
Heiko Schwarz,
Wojciech Samek
[pdf] [supp] [arXiv]
[bibtex]
@InProceedings{Becking_2022_CVPR, author = {Becking, Daniel and Kirchhoffer, Heiner and Tech, Gerhard and Haase, Paul and M\"uller, Karsten and Schwarz, Heiko and Samek, Wojciech}, title = {Adaptive Differential Filters for Fast and Communication-Efficient Federated Learning}, booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) Workshops}, month = {June}, year = {2022}, pages = {3367-3376} }

MPAF: Model Poisoning Attacks to Federated Learning Based on Fake Clients
Xiaoyu Cao,
Neil Zhenqiang Gong
[pdf] [arXiv]
[bibtex]
@InProceedings{Cao_2022_CVPR, author = {Cao, Xiaoyu and Gong, Neil Zhenqiang}, title = {MPAF: Model Poisoning Attacks to Federated Learning Based on Fake Clients}, booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) Workshops}, month = {June}, year = {2022}, pages = {3396-3404} }

Communication-Efficient Federated Data Augmentation on Non-IID Data
Hui Wen,
Yue Wu,
Jingjing Li,
Hancong Duan
[pdf]
[bibtex]
@InProceedings{Wen_2022_CVPR, author = {Wen, Hui and Wu, Yue and Li, Jingjing and Duan, Hancong}, title = {Communication-Efficient Federated Data Augmentation on Non-IID Data}, booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) Workshops}, month = {June}, year = {2022}, pages = {3377-3386} }

Does Federated Dropout Actually Work?
Gary Cheng,
Zachary Charles,
Zachary Garrett,
Keith Rush
[pdf]
[bibtex]
@InProceedings{Cheng_2022_CVPR, author = {Cheng, Gary and Charles, Zachary and Garrett, Zachary and Rush, Keith}, title = {Does Federated Dropout Actually Work?}, booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) Workshops}, month = {June}, year = {2022}, pages = {3387-3395} }