Adversarial Example

Adversarial examples are subtly altered inputs designed to fool machine learning models, primarily deep neural networks (DNNs), into making incorrect predictions. Current research focuses on improving model robustness against these attacks, exploring techniques like ensemble methods, multi-objective representation learning, and adversarial training, often applied to architectures such as ResNets and Vision Transformers. Understanding and mitigating the threat of adversarial examples is crucial for ensuring the reliability and security of AI systems across diverse applications, from image classification and natural language processing to malware detection and autonomous driving. The development of robust defenses and effective attack detection methods remains a significant area of ongoing investigation.

Papers

July 4, 2022

RAF: Recursive Adversarial Attacks on Face Recognition Using Extremely Limited Queries
Keshav Kasichainula, Hadi Mansourifar, Weidong Shi
Adversarial Attack Adversarial Example Face Recognition Successful Adversarial Attack

July 1, 2022

Efficient Adversarial Training With Data Pruning
Maximilian Kaufmann, Yiren Zhao, Ilia Shumailov, Robert Mullins, Nicolas Papernot
Adversarial Example Adversarial Training Adversarial Sample Data Pruning Adversarial Evaluation Efficient Adversarial Training

June 30, 2022

June 28, 2022

June 27, 2022

Adversarial Example Detection in Deployed Tree Ensembles
Laurens Devos, Wannes Meert, Jesse Davis
Adversarial Example Ensemble Learning Tree Ensemble Adversarial Detection Adversarial Example Detection

June 25, 2022

June 24, 2022

June 23, 2022

InfoAT: Improving Adversarial Training Using the Information Bottleneck Principle
Mengting Xu, Tao Zhang, Zhongnian Li, Daoqiang Zhang
Adversarial Example Adversarial Training Mutual Information Information Bottleneck

June 22, 2022

June 21, 2022

Transferable Graph Backdoor Attack
Shuiqiao Yang, Bao Gia Doan, Paul Montague, Olivier De Vel, Tamas Abraham, Seyit Camtepe, Damith C. Ranasinghe, Salil S. Kanhere
Graph Neural Network Adversarial Attack Adversarial Example Graph Backdoor Attack Transferable Backdoor Attack

June 20, 2022

Diversified Adversarial Attacks based on Conjugate Gradient Method
Keiichiro Yamamura, Haruki Sato, Nariaki Tateiwa, Nozomi Hata, Toru Mitsutake, Issa Oe, Hiroki Ishikura, Katsuki Fujisawa
Adversarial Attack Adversarial Example Conjugate Gradient Attack Algorithm

June 19, 2022

June 18, 2022

On the Role of Generalization in Transferability of Adversarial Examples
Yilin Wang, Farzan Farnia
Adversarial Attack Strong Generalization Adversarial Example Integral Role Task Transferability Generalization Property

Adversarial Example

Papers

RAF: Recursive Adversarial Attacks on Face Recognition Using Extremely Limited Queries

Efficient Adversarial Training With Data Pruning

MEAD: A Multi-Armed Approach for Evaluation of Adversarial Examples Detectors

Detecting and Recovering Adversarial Examples from Extracting Non-robust and Highly Predictive Adversarial Perturbations

Rethinking Adversarial Examples for Location Privacy Protection

Increasing Confidence in Adversarial Robustness Evaluations

Adversarial Example Detection in Deployed Tree Ensembles

Empirical Evaluation of Physical Adversarial Patch Attacks Against Overhead Object Detection Models

Defense against adversarial attacks on deep convolutional neural networks through nonlocal denoising

RSTAM: An Effective Black-Box Impersonation Attack on Face Recognition using a Mobile and Compact Printer

Defending Backdoor Attacks on Vision Transformer via Patch Processing

AdAUC: End-to-end Adversarial AUC Optimization Against Long-tail Problems

InfoAT: Improving Adversarial Training Using the Information Bottleneck Principle

AdvSmo: Black-box Adversarial Attack by Smoothing Linear Structure of Texture

SSMI: How to Make Objects of Interest Disappear without Accessing Object Detectors?

Transferable Graph Backdoor Attack

Diversified Adversarial Attacks based on Conjugate Gradient Method

On the Limitations of Stochastic Pre-processing Defenses

A Universal Adversarial Policy for Text Classifiers

On the Role of Generalization in Transferability of Adversarial Examples