paper-with-me

Papers

Universal Backdoor Attacks Detection via Adaptive Adversarial Probe

2022-09-12 · Yuhang Wang, Huafeng Shi, Rui Min, Ruijia Wu, Siyuan Liang, Yichao Wu, Ding Liang, Aishan Liu

Extensive evidence has demonstrated that deep neural networks (DNNs) are vulnerable to backdoor attacks, which motivates the development of backdoor attacks detection. Most detection methods are designed to verify whether a model is infected with presumed types of backdoor attacks, yet the adversary is likely to generate diverse backdoor attacks in practice that are unforeseen to defenders, which challenge current detection strategies. In this paper, we focus on this more challenging scenario and propose a universal backdoor attacks detection method named Adaptive Adversarial Probe (A2P). Specifically, we posit that the challenge of universal backdoor attacks detection lies in the fact that different backdoor attacks often exhibit diverse characteristics in trigger patterns (i.e., sizes and transparencies). Therefore, our A2P adopts a global-to-local probing framework, which adversarially probes images with adaptive regions/budgets to fit various backdoor triggers of different sizes/transparencies. Regarding the probing region, we propose the attention-guided region generation strategy that generates region proposals with different sizes/locations based on the attention of the target model, since trigger regions often manifest higher model activation. Considering the attack budget, we introduce the box-to-sparsity scheduling that iteratively increases the perturbation budget from box to sparse constraint, so that we could better activate different latent backdoors with different transparencies. Extensive experiments on multiple datasets (CIFAR-10, GTSRB, Tiny-ImageNet) demonstrate that our method outperforms state-of-the-art baselines by large margins (+12%).

📄 PDF Abstract BibTeX arXiv:2209.05244

Code (0)

등록된 구현이 없습니다.

Tasks

Scheduling

Similar Papers 제목 키워드 기반

Universal Soldier: Using Universal Adversarial Perturbations for Detecting Backdoor Attacks

2023-02-01 · Xiaoyun Xu, Oguzhan Ersoy, Stjepan Picek

Deep learning models achieve excellent performance in numerous machine learning tasks. Yet, they suffer from security-related issues such as adversarial examples and poisoning (backdoor) attacks. A deep learning model ma…

Adversarial Attack

Universal Litmus Patterns: Revealing Backdoor Attacks in CNNs

2019-06-26 · CVPR 2020 6 · Soheil Kolouri, Aniruddha Saha, Hamed Pirsiavash, Heiko Hoffmann

The unprecedented success of deep neural networks in many applications has made these networks a prime target for adversarial exploitation. In this paper, we introduce a benchmark technique for detecting backdoor attacks…

Traffic Sign Recognition

Contributor-Aware Defenses Against Adversarial Backdoor Attacks

2022-05-28 · Glenn Dawson, Muhammad Umer, Robi Polikar

Deep neural networks for image classification are well-known to be vulnerable to adversarial attacks. One such attack that has garnered recent attention is the adversarial backdoor attack, which has demonstrated the capa…

Backdoor Attackimage-classificationImage Classification

Towards Backdoor Attacks and Defense in Robust Machine Learning Models

2020-02-25 · Ezekiel Soremekun, Sakshi Udeshi, Sudipta Chattopadhyay

The introduction of robust optimisation has pushed the state-of-the-art in defending against adversarial attacks. Notably, the state-of-the-art projected gradient descent (PGD)-based training method has been shown to be …

BIG-bench Machine LearningClustering

Test-Time Backdoor Attacks on Multimodal Large Language Models

2024-02-13 · Dong Lu, Tianyu Pang, Chao Du, Qian Liu 외

Backdoor attacks are commonly executed by contaminating training data, such that a trigger can activate predetermined harmful effects during the test phase. In this work, we present AnyDoor, a test-time backdoor attack a…

Backdoor Attack