How to Craft Backdoors with Unlabeled Data Alone?

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Wang, Yifei, Ma, Wenhan, Jegelka, Stefanie, Wang, Yisen
Natura: Preprint
Pubblicazione: 2024
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866914765379469312
author Wang, Yifei
Ma, Wenhan
Jegelka, Stefanie
Wang, Yisen
author_facet Wang, Yifei
Ma, Wenhan
Jegelka, Stefanie
Wang, Yisen
contents Relying only on unlabeled data, Self-supervised learning (SSL) can learn rich features in an economical and scalable way. As the drive-horse for building foundation models, SSL has received a lot of attention recently with wide applications, which also raises security concerns where backdoor attack is a major type of threat: if the released dataset is maliciously poisoned, backdoored SSL models can behave badly when triggers are injected to test samples. The goal of this work is to investigate this potential risk. We notice that existing backdoors all require a considerable amount of \emph{labeled} data that may not be available for SSL. To circumvent this limitation, we explore a more restrictive setting called no-label backdoors, where we only have access to the unlabeled data alone, where the key challenge is how to select the proper poison set without using label information. We propose two strategies for poison selection: clustering-based selection using pseudolabels, and contrastive selection derived from the mutual information principle. Experiments on CIFAR-10 and ImageNet-100 show that both no-label backdoors are effective on many SSL methods and outperform random poisoning by a large margin. Code will be available at https://github.com/PKU-ML/nlb.
format Preprint
id arxiv_https___arxiv_org_abs_2404_06694
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle How to Craft Backdoors with Unlabeled Data Alone?
Wang, Yifei
Ma, Wenhan
Jegelka, Stefanie
Wang, Yisen
Machine Learning
Artificial Intelligence
Cryptography and Security
Relying only on unlabeled data, Self-supervised learning (SSL) can learn rich features in an economical and scalable way. As the drive-horse for building foundation models, SSL has received a lot of attention recently with wide applications, which also raises security concerns where backdoor attack is a major type of threat: if the released dataset is maliciously poisoned, backdoored SSL models can behave badly when triggers are injected to test samples. The goal of this work is to investigate this potential risk. We notice that existing backdoors all require a considerable amount of \emph{labeled} data that may not be available for SSL. To circumvent this limitation, we explore a more restrictive setting called no-label backdoors, where we only have access to the unlabeled data alone, where the key challenge is how to select the proper poison set without using label information. We propose two strategies for poison selection: clustering-based selection using pseudolabels, and contrastive selection derived from the mutual information principle. Experiments on CIFAR-10 and ImageNet-100 show that both no-label backdoors are effective on many SSL methods and outperform random poisoning by a large margin. Code will be available at https://github.com/PKU-ML/nlb.
title How to Craft Backdoors with Unlabeled Data Alone?
topic Machine Learning
Artificial Intelligence
Cryptography and Security
url https://arxiv.org/abs/2404.06694