First International Workshop on Federated Learning for Computer Vision (FedVision)

Federated Learning (FL) has become an important privacy-preserving paradigm in various machine learning tasks. However, the potential of FL in computer vision applications, such as face recognition, person re-identification, and action recognition, is far from being fully exploited. Moreover, FL has rarely been demonstrated effectively in advanced computer vision tasks such as object detection and image segmentation, compared to the traditional centralized training paradigm.

This workshop aims at bringing together researchers and practitioners with common interests in FL for computer vision and studying the different synergistic relations in this interdisciplinary area. The day-long event will facilitate interaction among students, scholars, and industry professionals from around the world to discuss future research challenges and opportunities.

Keynote Speakers

Program Schedule (Time in CDT)

8:30 AM – 8:40 AM (CDT) Chairs’ opening remarks [Slide] [Video]

8:40 AM – 9:25 AM Keynote talk 1 (in person)

"Privacy-preserving sensors for Encrypted Visual AI", Dr. Jeffrey Byrne (Visym Labs) [Slide] [Video]

9:30 AM – 10:15 AM Keynote talk 2 (online)

"OpenFL: A framework for federated learning", Dr. Brandon Edwards (Intel) [Slide] [Video]

10:15 AM – 10:30 AM Break

10:30 AM – 11:15 AM Keynote talk 3 (in person)

"Scalable, Heterogeneity-Aware and Trustworthy Federated Learning", Dr. Yiran Chen (Duke University) [Slide] [Video]

11:20 AM – 12:05 PM Keynote talk 4 (in person)

"Federated Knowledge Composition", Dr. Handong Zhao (Adobe) [Slide] [Video]

12:05 PM – 1:30 PM Lunch break

1:30PM – 2:15 PM Keynote talk 5 (online)

"Social, Secure, Scalable, and Efficient Federated Learning", Dr. Salman Avestimehr (USC) [Slide] [Video]

2:20 PM – 3:05 PM Keynote talk 6 (online)

"Coping with Heterogeneity and Privacy in Communication-Efficient Federated Optimization", Dr. Yuejie Chi (CMU) [Slide] [Video]

3:05 PM – 3:15 PM Break

3:15 PM – 4:00 PM Keynote talk 7 (online)

"Cross-device FL: optimization and privacy", Dr. Zheng Xu (Google) [Slide] [Video]

4:00 PM – 5:00 PM Oral session

Paper #1 (4:10 - 4:20 PM):

FedIris: Towards More Accurate and Privacy-preserving Iris Recognition via Federated Template Communication [Slide] [Video]

Zhengquan Luo ( University of Science and Technology of China(USTC), Center for Research on Intelligent Perception and Computing (CRIPAC), Institute of Automation, Chinese Academy of Sciences (CASIA) ); Yunlong Wang ( Center for Research on Intelligent Perception and Computing (CRIPAC) National Laboratory of Pattern Recognition (NLPR) Institute of Automation, Chinese Academy of Sciences (CASIA) ); Zilei Wang ( University of Science and Technology of China ); Zhenan Sun ( Chinese of Academy of Sciences ); Tieniu Tan ( NLPR, China )

Paper #2 (4:20 - 4:30 PM):

Adaptive Differential Filters for Fast and Communication-Efficient Federated Learning [Slide] [Video]

Daniel Becking ( Fraunhofer HHI ); Heiner Kirchhoffer ( Fraunhofer HHI ); Gerhard Tech ( Fraunhofer Heinrich-Hertz Institute ); Paul Haase ( Fraunhofer HHI ); Karsten Müller ( Heinrich Hertz Institute ); Heiko Schwarz ( Fraunhofer-Institut für Nachrichtentechnik, Heinrich-Hertz-Institut, HHI ); Wojciech Samek ( Fraunhofer HHI )

Paper #3 (4:30 - 4:40 PM):

Communication-Efficient Federated Data Augmentation on Non-IID Data [Slide] [Video]

Hui Wen ( University of Electronic Science and Technology of China ); Yue Wu ( University of Electronic Science and Technology of China ); Jingjing Li ( University of Electronic Science and Technology of China ) ; Hancong Duan ( University of Electronic Science and Technology of China )

Paper #4 (4:40 - 4:50 PM):

Does Federated Dropout actually work? [Video]

Gary Cheng ( Stanford University ); Zachary Charles ( Google Research ); Zachary Garrett ( Google Research ); John Rush ( Google Research )

Paper #5 (4:50 - 5:00 PM):

MPAF: Model Poisoning Attacks to Federated Learning based on Fake Clients [Slide] [Video]

Xiaoyu Cao ( Duke University ); Neil Zhenqiang Gong ( Duke University )