Work

Understanding Deep Gradient Leakage via Inversion Influence Functions

Year: 2023

Type: article

Abstract: F effectively approximated the DGL generally on different model architectures, datasets, modalities, attack implementations, and perturbation-based defenses. With this novel tool, we provide insights into effective gradient perturbation directions, the unfairness of privacy protection, and privacy-preferred model initialization. Our codes are provided in https://github.com/illidanlab/inversion-influence-function.

Source: PubMed

DOI: 10.48550/arxiv.2309.13016

Authors: Zhang, Haobo, Hong, Junyuan, Deng, Yuyang, Mahdavi, Mehrdad, Zhou, Jiayu

Institutions: Michigan State University, The University of Texas at Austin, Pennsylvania State University

Language: English

Topic: Privacy-Preserving Technologies in Data

Subfield: Artificial Intelligence

Field: Computer Science

Domain: Physical Sciences

Open Access status: green

Funders: National Science Foundation, National Institute on Aging, Office of Naval Research