How Does a Deep Learning Model Architecture Impact Its Privacy? A Comprehensive Study of Privacy Attacks on CNNs and Transformers

Authors: 

Guangsheng Zhang, Bo Liu, Huan Tian, and Tianqing Zhu, University of Technology Sydney; Ming Ding, Data 61, Australia; Wanlei Zhou, City University of Macau

Abstract: 

As a booming research area in the past decade, deep learning technologies have been driven by big data collected and processed on an unprecedented scale. However, privacy concerns arise due to the potential leakage of sensitive information from the training data. Recent research has revealed that deep learning models are vulnerable to various privacy attacks, including membership inference attacks, attribute inference attacks, and gradient inversion attacks. Notably, the efficacy of these attacks varies from model to model. In this paper, we answer a fundamental question: Does model architecture affect model privacy? By investigating representative model architectures from convolutional neural networks (CNNs) to Transformers, we demonstrate that Transformers generally exhibit higher vulnerability to privacy attacks than CNNs. Additionally, we identify the micro design of activation layers, stem layers, and LN layers, as major factors contributing to the resilience of CNNs against privacy attacks, while the presence of attention modules is another main factor that exacerbates the privacy vulnerability of Transformers. Our discovery reveals valuable insights for deep learning models to defend against privacy attacks and inspires the research community to develop privacy-friendly model architectures.

Open Access Media

USENIX is committed to Open Access to the research presented at our events. Papers and proceedings are freely available to everyone once the event begins. Any video, audio, and/or slides that are posted after the event are also free and open to everyone. Support USENIX and our commitment to Open Access.

BibTeX
@inproceedings {298044,
author = {Guangsheng Zhang and Bo Liu and Huan Tian and Tianqing Zhu and Ming Ding and Wanlei Zhou},
title = {How Does a Deep Learning Model Architecture Impact Its Privacy? A Comprehensive Study of Privacy Attacks on {CNNs} and Transformers},
booktitle = {33rd USENIX Security Symposium (USENIX Security 24)},
year = {2024},
isbn = {978-1-939133-44-1},
address = {Philadelphia, PA},
pages = {6795--6812},
url = {https://www.usenix.org/conference/usenixsecurity24/presentation/zhang-guangsheng},
publisher = {USENIX Association},
month = aug
}