English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Federated learning (FL) is a popular distributed learning framework that canreduce privacy risks by not explicitly sharing private data. However, recentworks demonstrated that sharing model updates makes FL vulnerable to inferenceattacks. In this work, we show our key observation that the data representationleakage from gradients is the essential cause of privacy leakage in FL. We alsoprovide an analysis of this observation to explain how the data presentation isleaked. Based on this observation, we propose a defense against model inversionattack in FL. The key idea of our defense is learning to perturb datarepresentation such that the quality of the reconstructed data is severelydegraded, while FL performance is maintained. In addition, we derive certifiedrobustness guarantee to FL and convergence guarantee to FedAvg, after applyingour defense. To evaluate our defense, we conduct experiments on MNIST andCIFAR10 for defending against the DLG attack and GS attack. Without sacrificingaccuracy, the results demonstrate that our proposed defense can increase themean squared error between the reconstructed data and the raw data by as muchas more than 160X for both DLG attack and GS attack, compared with baselinedefense methods. The privacy of the FL system is significantly improved.

Provable Defense against Privacy Leakage in Federated Learning from Representation Perspective