FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W4386807902> ?p ?o ?g. }

• W4386807902 abstract "Federated Learning (FL) enables distributed participants (e.g., mobile devices) to train a global model without sharing data directly to a central server. Recent studies have revealed that FL is vulnerable to gradient inversion attack (GIA), which aims to reconstruct the original training samples and poses high risk against the privacy of clients in FL. However, most existing GIAs necessitate control over the server and rely on strong prior knowledge including batch normalization and data distribution information. In this work, we propose Client-side poisoning Gradient Inversion (CGI), which is a novel attack method that can be launched from clients. For the first time, we show the feasibility of a client-side adversary with limited knowledge being able to recover the training samples from the aggregated global model. We take a distinct approach in which the adversary utilizes a malicious model that amplifies the loss of a specific targeted class of interest. When honest clients employ the poisoned global model, the gradients of samples belonging to the targeted class are magnified, making them the dominant factor in the aggregated update. This enables the adversary to effectively reconstruct the private input belonging to other clients using the aggregated update. In addition, our CGI also features its ability to remain stealthy against Byzantine-robust aggregation rules (AGRs). By optimizing malicious updates and blending benign updates with a malicious replacement vector, our method remains undetected by these defense mechanisms. To evaluate the performance of CGI, we conduct experiments on various benchmark datasets, considering representative Byzantine-robust AGRs, and exploring diverse FL settings with different levels of adversary knowledge about the data. Our results demonstrate that CGI consistently and successfully extracts training input in all tested scenarios." @default.
• W4386807902 created "2023-09-17" @default.
• W4386807902 creator A5008056593 @default.
• W4386807902 creator A5009580845 @default.
• W4386807902 creator A5013793366 @default.
• W4386807902 creator A5015011245 @default.
• W4386807902 creator A5034348483 @default.
• W4386807902 creator A5049263948 @default.
• W4386807902 creator A5051439492 @default.
• W4386807902 creator A5056248574 @default.
• W4386807902 date "2023-09-13" @default.
• W4386807902 modified "2023-10-16" @default.
• W4386807902 title "Client-side Gradient Inversion Against Federated Learning from Poisoning" @default.
• W4386807902 doi "https://doi.org/10.48550/arxiv.2309.07415" @default.
• W4386807902 hasPublicationYear "2023" @default.
• W4386807902 type Work @default.
• W4386807902 citedByCount "0" @default.
• W4386807902 crossrefType "posted-content" @default.
• W4386807902 hasAuthorship W4386807902A5008056593 @default.
• W4386807902 hasAuthorship W4386807902A5009580845 @default.
• W4386807902 hasAuthorship W4386807902A5013793366 @default.
• W4386807902 hasAuthorship W4386807902A5015011245 @default.
• W4386807902 hasAuthorship W4386807902A5034348483 @default.
• W4386807902 hasAuthorship W4386807902A5049263948 @default.
• W4386807902 hasAuthorship W4386807902A5051439492 @default.
• W4386807902 hasAuthorship W4386807902A5056248574 @default.
• W4386807902 hasBestOaLocation W43868079021 @default.
• W4386807902 hasConcept C109007969 @default.
• W4386807902 hasConcept C119857082 @default.
• W4386807902 hasConcept C120314980 @default.
• W4386807902 hasConcept C13280743 @default.
• W4386807902 hasConcept C136886441 @default.
• W4386807902 hasConcept C144024400 @default.
• W4386807902 hasConcept C151730666 @default.
• W4386807902 hasConcept C154945302 @default.
• W4386807902 hasConcept C185798385 @default.
• W4386807902 hasConcept C1893757 @default.
• W4386807902 hasConcept C19165224 @default.
• W4386807902 hasConcept C202477664 @default.
• W4386807902 hasConcept C205649164 @default.
• W4386807902 hasConcept C2992525071 @default.
• W4386807902 hasConcept C31258907 @default.
• W4386807902 hasConcept C37736160 @default.
• W4386807902 hasConcept C38652104 @default.
• W4386807902 hasConcept C41008148 @default.
• W4386807902 hasConcept C41065033 @default.
• W4386807902 hasConcept C86803240 @default.
• W4386807902 hasConceptScore W4386807902C109007969 @default.
• W4386807902 hasConceptScore W4386807902C119857082 @default.
• W4386807902 hasConceptScore W4386807902C120314980 @default.
• W4386807902 hasConceptScore W4386807902C13280743 @default.
• W4386807902 hasConceptScore W4386807902C136886441 @default.
• W4386807902 hasConceptScore W4386807902C144024400 @default.
• W4386807902 hasConceptScore W4386807902C151730666 @default.
• W4386807902 hasConceptScore W4386807902C154945302 @default.
• W4386807902 hasConceptScore W4386807902C185798385 @default.
• W4386807902 hasConceptScore W4386807902C1893757 @default.
• W4386807902 hasConceptScore W4386807902C19165224 @default.
• W4386807902 hasConceptScore W4386807902C202477664 @default.
• W4386807902 hasConceptScore W4386807902C205649164 @default.
• W4386807902 hasConceptScore W4386807902C2992525071 @default.
• W4386807902 hasConceptScore W4386807902C31258907 @default.
• W4386807902 hasConceptScore W4386807902C37736160 @default.
• W4386807902 hasConceptScore W4386807902C38652104 @default.
• W4386807902 hasConceptScore W4386807902C41008148 @default.
• W4386807902 hasConceptScore W4386807902C41065033 @default.
• W4386807902 hasConceptScore W4386807902C86803240 @default.
• W4386807902 hasLocation W43868079021 @default.
• W4386807902 hasOpenAccess W4386807902 @default.
• W4386807902 hasPrimaryLocation W43868079021 @default.
• W4386807902 hasRelatedWork W2522301850 @default.
• W4386807902 hasRelatedWork W3046843850 @default.
• W4386807902 hasRelatedWork W3127875616 @default.
• W4386807902 hasRelatedWork W3157496929 @default.
• W4386807902 hasRelatedWork W3200471468 @default.
• W4386807902 hasRelatedWork W4248052496 @default.
• W4386807902 hasRelatedWork W4251088474 @default.
• W4386807902 hasRelatedWork W43121067 @default.
• W4386807902 hasRelatedWork W4362599004 @default.
• W4386807902 hasRelatedWork W4386716251 @default.
• W4386807902 isParatext "false" @default.
• W4386807902 isRetracted "false" @default.
• W4386807902 workType "article" @default.