Diagnosing Vulnerability of Variational Auto-Encoders to Adversarial Attacks

In this work, we explore adversarial attacks on the Variational Autoencoders (VAE). We show how to modify data point to obtain a prescribed latent code (supervised attack) or just get a drastically different code (unsupervised attack). We examine the influence of model modifications ($β$-VAE, NVAE) on the robustness of VAEs and suggest metrics to quantify it.

cs.CR cs.LG stat.ML Adversarial Robustness in Machine Learning Generative Adversarial Networks and Image Synthesis Digital Media Forensic Detection Computer Science Mathematics

Paper

Similar papers

Adversarial Attacks on Variational AutoencodersGeorge Gondim-Ribeiro, Pedro Tabacof, Eduardo Valle2018 · arXiv (Cornell University) · 46 citations In Library

Adversarial Attacks on Variational AutoencodersPedro Tabacof2018 In Library

Alleviating Adversarial Attacks on Variational Autoencoders with MCMCAnna Kuzina, Max Welling, Jakub M. Tomczak2022 · arXiv (Cornell University) · 16 citations In Library

On the Adversarial Robustness of Generative Autoencoders in the Latent SpaceMingfei Lu, Badong Chen2023 · arXiv (Cornell University) · 7 citations In Library

Adversarial robustness of VAEs through the lens of local geometryAsif Khan, Amos Storkey2022 · arXiv (Cornell University) · 1 citation In Library

Adversarial Images for Variational AutoencodersPedro Tabacof, Julia Tavares, Eduardo Valle2016 · arXiv.org · 92 citations In Library

Certifiably Robust Variational AutoencodersBen Barrett, Alexander Camuto, Matthew Willetts et al.2021 · arXiv (Cornell University) · 18 citations In Library

Towards a Theoretical Understanding of the Robustness of Variational AutoencodersAlexander Camuto, Matthew Willetts, Stephen Roberts et al.2020 · arXiv (Cornell University) · 34 citations In Library

Learning from Attacks: Attacking Variational Autoencoder for Improving Image ClassificationZheng, Jianzhang, Yang, Fan, Hao Shen et al.2022 · arXiv (Cornell University) · 0 citations In Library

Adversarial Robustness of VAEs across Intersectional SubgroupsChethan Krishnamurthy Ramanaik, Arjun Roy, Eirini Ntoutsi2024 · arXiv (Cornell University) · 1 citation In Library

Source of record

arXiv

44B

Welcome.

This panel sticks with you. Pick anything below and it opens right beside it, so you can dig through 60,000-plus records without ever losing your spot here.

Travel 44B

LibraryPapers, policy, standards, statute LabsEvery organization building AI ModelsIntelligence, price, and speed BenchmarksThe evaluation catalog IncidentsAI failures and harms SearchOne field across everything DashboardThe state of AI accountability in NY 44B RegistryThe Art. 44-B compliance portal