Generating Private Data Surrogates for Vision Related Tasks

Webster, Ryan; Rabin, Julien; Simon, Loïc; Jurie, Frédéric

doi:10.1109/icpr48806.2021.9413067

Cited by 4 publications

(1 citation statement)

References 25 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Besides, following the work of [4] showing some form of GAN generalization, some attempts were made to leverage GANs to produce ersatz datasets aiming at more privacy. For instance in [18], a dataset, which is kept private, is used to train a GAN. Then the GAN is made publicly available for downstream tasks.…”

Section: Introductionmentioning

confidence: 99%

This Person (Probably) Exists. Identity Membership Attacks Against GAN Generated Faces

Webster,

Rabin,

Simon

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

Recently, generative adversarial networks (GANs) have achieved stunning realism, fooling even human observers. Indeed, the popular tongue-in-cheek website http://thispersondoesnotexist.com, taunts users with GAN generated images that seem too real to believe. On the other hand, GANs do leak information about their training data, as evidenced by membership attacks recently demonstrated in the literature. In this work, we challenge the assumption that GAN faces really are novel creations, by constructing a successful membership attack of a new kind. Unlike previous works, our attack can accurately discern samples sharing the same identity as training samples without being the same samples. We demonstrate the interest of our attack across several popular face datasets and GAN training procedures. Notably, we show that even in the presence of significant dataset diversity, an over represented person can pose a privacy concern.

show abstract

Section: Introductionmentioning

confidence: 99%

This Person (Probably) Exists. Identity Membership Attacks Against GAN Generated Faces

Webster,

Rabin,

Simon

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

A survey on membership inference attacks and defenses in machine learning

Niu,

Liu,

Zhu

et al. 2024

Journal of Information and Intelligence

View full text Add to dashboard Cite

Defenses to Membership Inference Attacks: A Survey

Hu,

Yan,

Yan

et al. 2023

ACM Comput. Surv.

View full text Add to dashboard Cite

Machine learning (ML) has gained widespread adoption in a variety of fields, including computer vision and natural language processing. However, ML models are vulnerable to membership inference attacks (MIAs), which can infer whether access data was used in training a target model, thus compromising the privacy of training data. This has led researchers to focus on protecting the privacy of ML. To date, although there have been extensive efforts to defend against MIAs, we still lack a comprehensive understanding of the progress made in this area, which can often impede our ability to design the most effective defense strategies. In this paper, we aim to fill this critical knowledge gap by providing a systematic analysis of membership inference defense.Specifically, we classify and summarize the existing membership inference defense schemes, focusing on optimization phase and objective, basic intuition, and key technology, and discuss possible research directions of membership inference defense in the future.

show abstract

Generating Private Data Surrogates for Vision Related Tasks

Cited by 4 publications

References 25 publications

This Person (Probably) Exists. Identity Membership Attacks Against GAN Generated Faces

This Person (Probably) Exists. Identity Membership Attacks Against GAN Generated Faces

A survey on membership inference attacks and defenses in machine learning

Defenses to Membership Inference Attacks: A Survey

Contact Info

Product

Resources

About