Adversarial Attack

Adversarial attacks aim to deceive machine learning models by subtly altering input data, causing misclassifications or other erroneous outputs. Current research focuses on developing more robust models and detection methods, exploring various attack strategies across different model architectures (including vision transformers, recurrent neural networks, and graph neural networks) and data types (images, text, signals, and tabular data). Understanding and mitigating these attacks is crucial for ensuring the reliability and security of AI systems in diverse applications, from autonomous vehicles to medical diagnosis and cybersecurity.

Papers

February 1, 2024

January 31, 2024

Adversarial Quantum Machine Learning: An Information-Theoretic Generalization Analysis
Petros Georgiou, Sharu Theresa Jose, Osvaldo Simeone
Adversarial Attack Quantum Classifier Information Theoretic Generalization

January 30, 2024

January 28, 2024

Addressing Noise and Efficiency Issues in Graph-Based Machine Learning Models From the Perspective of Adversarial Attack
Yongyu Wang
Adversarial Attack High Efficiency Visual Perspective Graph Based Learning Graph Construction Graph Algorithm Mitigating Noise

January 27, 2024

January 26, 2024

January 25, 2024

January 24, 2024

AdCorDA: Classifier Refinement via Adversarial Correction and Domain Adaptation
Lulan Shen, Ali Edalati, Brett Meyer, Warren Gross, James J. Clark
Adversarial Attack Domain Adaptation Adversarial Learning Training Set Training Based Model Refinement

January 23, 2024

January 22, 2024

January 21, 2024

Adversarial Attack

Papers

Benchmarking Transferable Adversarial Attacks

Invariance-powered Trustworthy Defense via Remove Then Restore

Adversarial Quantum Machine Learning: An Information-Theoretic Generalization Analysis

Robust Prompt Optimization for Defending Language Models Against Jailbreaking Attacks

Towards Assessing the Synthetic-to-Measured Adversarial Vulnerability of SAR ATR

Addressing Noise and Efficiency Issues in Graph-Based Machine Learning Models From the Perspective of Adversarial Attack

L-AutoDA: Leveraging Large Language Models for Automated Decision-based Adversarial Attacks

Asymptotic Behavior of Adversarial Training Estimator under $\ell_\infty$-Perturbation

Set-Based Training for Neural Network Verification

Conserve-Update-Revise to Cure Generalization and Robustness Trade-off in Adversarial Training

Friendly Attacks to Improve Channel Coding Reliability

Sparse and Transferable Universal Singular Vectors Attack

Respect the model: Fine-grained and Robust Explanation with Sharing Ratio Decomposition

AdCorDA: Classifier Refinement via Adversarial Correction and Domain Adaptation

The twin peaks of learning neural networks

DAFA: Distance-Aware Fair Adversarial Training

A Training-Free Defense Framework for Robust Learned Image Compression

Cloud-based XAI Services for Assessing Open Repository Models Under Adversarial Attacks

How Robust Are Energy-Based Models Trained With Equilibrium Propagation?

Finding a Needle in the Adversarial Haystack: A Targeted Paraphrasing Approach For Uncovering Edge Cases with Minimal Distribution Distortion